It does sound interesting experimental WS, DS and JS designs that never got off the drawing board. It would likely be a book full of designs like the Enterprise. Great start but unfortunately it got lost as it went along ultimately leading to its failure.
Something along those lines, but a "failed" design doesn't necessarily have to be a failure. We've been told that some designs were excepted for very specific reasons, like the number of DropShip collars, the engines, cost, etc. Who knows how many suitable designs were submitted to the procurement office, but never built?
That's the fun part about a book like this, you could mix and match truly "failed" designs with some that were very good, but ignored for one reason of another that happened to be pertinent at the time of submission.
If I was writing it and I wanted to make the book truly special, I'd run every design through MM against some canon vehicle—just like they'd do it at the academy. The results would form the basis for the "results" section in every entry.
If I had time, or if I could get MM to work on my Mac, I'd probably give this idea a go.
