Apologies for resuscitating a dead thread, but the logistics & economics of BT are one of the things I love to play with.
The original House Marik sourcebook specifically cites 488 mechs produced in the FWL per year; 141 light 'mechs, 187 mediums, 126 heavies and 34 assaults. These are produced at 13 different factory locations owned by 11 different companies. This info can also be broken down a few other ways that may or may not prove illuminating, depending on how much you care about logistics.
# of factories by weight class:
7 factories produce a total of four types of light 'mech
10 factories produce a total of nine types of medium 'mech (3 factories produce two designs)
9 factories produce a total of seven types of heavy 'mech (1 factory produces two designs)
4 factories produce a total of four types of assault 'mech
However, there are some internal errors in the book.
In the text, Irian BattleMechs is said to produce Quickdraws and Wasps on Shiro III, and Hermes IIs on Irian itself. In the table listing 'mechs produced by factory, however, the Shiro III plant is listed as building Wasps (31), Hermes IIs (20) and Stalkers (11), while the Irian plant builds Hermes IIs (14) and Awesomes (7). The table provides the 488 mechs/year and the 13 factories/11 companies statistics. But should we believe it over the text, or should the text trump the table? Can these be reconciled? Throw in the fact that the text also says the FWL has 17 "full or partially operational 'mech production facilities" (pg 114 for any interested), yet only lists 13. Three of those missing could be the StarCorps and Kallon lines of Emris IV, Bernardo and Asuncion (more on these below), but it seems more likely they're part of the seven "shut down (due to component shortages) or disabled" 'mech production lines also mentioned.
More amusingly, the text of the Marik book claims that due to the relatively few Assault mechs manufactured after the loss of Carver V's Atlas & Victor lines (see more below), the Phoenix Hawk and Hunchback are beginning to "dominate" the FWL medium lineup...even though the table lists the P-hawk as tied for least-produced-mediums (with the Vulcan, at 13 mechs/year) and the Huncback only slightly better (14/year, tied with the Shadow Hawk). At the same time, 41 Griffins are made each year, as are 36 Wolverine-Ms; so are 34 Hermes IIs, but they're also listed as dominating the medium weight-class alongside the P-hawk and H-back. Clearly it should be the Hermes II, Griffin, and Wolverine that are dominating the medium lances.
The Earthwerks corporate profile text says the factory on Keystone produces Stingers, Thunderbolts and Griffins; the 'mech production table says it produces Stingers, Griffins, Phoenix Hawks and BattleMasters. The company profile text goes on to say that the Calloway VI factory makes Stingers and Archers, with a third line retooling to make Phoenix Hawks (what they used to make is anyone's guess). But the table-of-production gives the Calloway VI plant Stingers, Archers and Shadow Hawks, making it hard to determine if that third (presumably Shadow Hawk) line is being converted, or if there's actually a fourth production line as at the Keystone facility.
Historical: Brush Wars claims there was an undamaged SL-era StarCorps 'mech factory on Emris IV in 3014, and provides no indication that it was destroyed. It mentions an event depicted in...Tales of the Black Widow, I believe, but I don't recall that comic depicting the StarCorps factory as destroyed, either. So either the factory should still be there and the original Marik book is in error or it needs to have been destroyed within 10 years of Anton's Revolt. Or my memory could be wrong and the TotBW comic shows the StarCorps factory in ruins after the Dragoons take it. But that's not the Dragoons' style, is it?
The original House Liao sourcebook, meanwhile, mentions Kallon factories on Bernardo and Asuncion, now producing 'mechs for the FWL, but no indication of them can be found in the House Marik book; Kallon only has one factory in the Marik book, on Thermopolis (which produces 18 Wolverine-Ms and 8 Riflemans per year), and neither Bernardo nor Asuncion are listed among Kallon's Wolverine or Rifleman factories in either the Marik book or TRO 3039; they seem to have fallen out of canon.
In the JumpShip & DropShip realm, the Marik sourcebook claims the FWL is the leader of spacecraft production in the Inner Sphere, as it produces a fantastic 3 JumpShips and 8 DropShips each year! Wow! "Gateway to the Stars" indeed! Not an internal error, as it's more or less in line with DropShips & JumpShips, which I recall claims only 12 Jumpships are produced each year in the entire Inner Sphere, but still amusing.
Liao production notes:
As for what's listed in the original Liao Handbook, the factory on Grand Base is interesting. The 3025 book claims it's new, Historical: Brush Wars seems to wreck much of it in 3032, and 20 Year Update specifically cites all four remaining Capellan factories needing to go up to six years without any production due to supply line disruptions caused by the 4th Succession War, with Grand Base being the last to resume production (in 3036).
Meanwhile, although TRO 3039 lists a Kallon factory on Nanking (with Wolverines and Riflemen produced there), and although Nanking is cited as the Headquarters of Kallon's operations in their entry in the 3025 Liao sourcebook, Nanking is not among the 8 factory-planets listed in that same Liao book. Clearly it's producing for the FedSuns by 3039, however. Either it got rebuilt quite quickly by the nascent FedCom economy, or else the Liaos should have had more than the 400 'mechs cited in the Liao book.
The Marik handbook claims there was an Atlas & Victor production line on Carver V, recovered by Liao in 2968. It must have been subsequently destroyed, however, because such a factory is not listed in the corresponding Liao book.
I haven't gone through the Davion book for details yet, largely because I know it doesn't throw them where anyone can find them the way the Liao and Marik books did. And the sourcebooks, as always, are sufficiently prone to error to make it annoying when these inconsistencies keep cropping up.
Conclusions:
The Marik book more than once mentions that the FWL is falling behind the other states, militarily. The same is echoed in the Liao book, which vividly describes the Capellan economy (and even the military-industrial complex) as falling apart. Therefore, the 488+ mechs per year of the FWL production and 400+ mechs/year of Capellan production should be taken as all that the least militarily-active Successor States can produce; the massive Lyran economy and the militarily very-active Federated Suns and Draconis Combine should be producing significantly more BattleMechs per year in order to keep their larger militaries in better shape.
Between the Federal & Provincial forces, the FWL fields 55 BattleMech regiments in 3025 (not including 2 regiments & 7 battalions of mercenaries). This means that each FWL regiment gets 8.87 new BattleMechs per year, assuming all production from within the FWL stays within the FWL, and assuming no 'mechs are imported. That's not a lot.
The Liao, meanwhile, have 127 battalions, not including mercenaries (per the original Liao handbook's deployment chart, and including the Confederation Reserves, which are listed on the deployment chart but not provided with an entry in the Unit List section); if there were 3 battalions per regiment, this would come out to 42 regiments and one additional battalion (it's actually 44 regiments, the Death Commandos, and one 'mech battalion for each of the 8 Warrior Houses). Or 9.45 mechs per regiment per year (3.15 per battalion per year). Which is, in fact, a slightly better replacement ratio than the Free Worlds. But as 60 of the Liao 'mechs produced each year are from the new (in 3025!) factory on Grand Base, before that facility opened the Liao were doing much worse than the Leaguers, at 8.03 mechs/regiment/year (2.68 mechs/battalion/year).
If these two nations are providing a fair number 'mechs to their mercenary commands (which would certainly be the case for the MAC), then their ability to rebuild shattered forces (such as after Anton's Revolt, the Andurien Succession & Andurien/Canopian war against the Capellans, etc) should be quite limited. For example, while Thomas was building the Free Worlds Legionnaires, with four regiments fielded in three years immediately following the Andurien crisis, diverting so many 'mechs to those units (and continuing to build another four regiments in the following ten years) should have greatly weakened all other 'mech forces in the FWL, particularly those loyalist units which had taken a beating in the conquest of Andurien. This diversion of resources should have quite upset the units which bled to deal with the Andurien succession...unless the FWL somehow managed to drastically increase its production facilities while the Andurien revolt was on-going (which seems unlikely).
Six regiments destroyed in Anton's revolt were reconstituted by 3025; that alone would require nearly 18 months of 'mech production, not including the dozens of other units heavily damaged during the revolt, which seemingly caused the Orloff Grenadiers to cannabilize the 5th regiment to bring the 1st, 6th and 8th regiments back up to strength. One regiment of the Oriente Hussars was destroyed and not rebuilt. The Regulan Hussars had one regiment destroyed (but rebuilt it within a decade) and three savagely mauled. Hopefully Janos Marik figured out where Anton acquired four regiments' worth of 'mechs to create the first four Ducal Guard regiments, and used that as a source of equipment to rebuild the 6th, 9th, and 18th Marik Militias, along with the 1st and 12th Atrean Dragoons (the 11th at least still had some mechs survive the conflict and so only needed partial rebuilding). And the post-war reconstruction would not be able to be used to replace losses from engagements with the Lyrans or Capellans.
Tracking how badly the CCAF was mauled in the 4th War is rather trickier, because the 4th SW books don't provide handy tables, but if there's sufficient interest I'd be happy to look through those for a detailed look, but I'm quite impressed that the Capellans managed to bring all five MAC regiments back online by 3032, given that at least one of them was destroyed wholesale. Then again, NAIS 4th SW vol 2 says the Capellans ended the war with only 41 'mech battalions, but Historical: Brush Wars lists them with 57 in 3032, despite not having any 'mech production on-line at that time (per 20 Year Update), and having been forced to abandon their wrecked-but-salvageable 'mechs on pretty much every world in the 4th SW (and thus not being able to reconstitute their forces through that usual method).
At a guess, the Lyrans are probably building in the nature of 6-800 mechs/year, even though they only have 52 House regiments in 3025; their economy is quite strong, and their military commanders quite stupid, so they'll need replacement 'mechs more frequently in order to retain what LCAF units they still retain, and to supply to some of the ~23 mercenary regiments they employ. At twice the replacement rate of the FWL forces (to account for Lyran incompetence) this would come out to about 925 'mechs per year, which is probably rather much, but it'd be hard to argue it below about 700 mechs/year for 13.46 mechs/regiment/year for the 52 LCAF units.
The AFFS might build fewer 'mechs to support its 73 regiments/221 battalions (including March Militias but not mercenaries), because its leaders are slightly more competent. On the other hand, they still have difficulties like the Galtor campaign which would require rebuilding whole units, and they employ 28 mercenary regiments/86 mercenary battalions. At Marik or Liao replacement rates, this would come out to around 650-700 'mechs per year.
The Kuritans deployed 62 regiments for 193 battalions (not including mercs), so even at Marik or Liao replacement levels they'd be building 550-600 'mechs per year.