Mark Wade at www.astronautix.com does have the editorial position that even given the “modular capsule” concept, CEV as decided upon is a triumph of political compromise over engineering and NASA forewent better private-industry designs, to a great extent due to budgetary and political-timetable pressure to get something flying fast, and if possible using the same workforces that are already involved in the Shuttle program, plus a bit of “not invented here” syndrome. I think he’s a bit pessimistic, though the decision on the configuration does lose some flexibility (Wade, and myself, would have preferred something more along the Soyuz philosophy of 3 modules: systems/propulsion; command/reentry; operations/docking – with the latter being suitable for “stretching” in order to do different mission profiles.)
Of course, part of the problem here is that Manned Space Flight is not a profitmaker, so while Lockheed-Martin, Boeing, etc. may have excellent ideas as to how to build a spacecraft right, they have no excuse whatsoever to sink real money into building one unless they already have in their hand the contract from NASA. So they instead have to compete on the basis of who will meet NASA proposal specs, whatever they are, which may or may not be what’s better. But then reality collides with NASA proposal specs and everything has to be revised and, if you’re lucky, someone at some point says, “enough! build what we have worked out so far!” and you end up with a compromise system – or more often, your project just gets cancelled after pouring heavy millions in just design and testing(Remember Space Station Freedom?). And of course, you then have to oversell capabilities and lowball price in order to get the Appropriations Committee to even consider giving you a budget (see: Space Shuttle)
K364, actually, the STS looks to me more like a 1.5-stage vehicle (there are the boosters and SMEs, but also orbital insertion engines that are internally fueled).
The last Shuttle, Endeavour, is payload rated to 24 metric tonnes into LEO, more than an Atlas-V, way less than an Ariane-V. An STS can fly operationally with a crew as small as 3 (pilot, copilot, mission specialist to handle the experiment/satellite launch) or as large as 8.
The real mistake with the shuttle was the attempt at having it be the one do-everything MSF system AND heavy launch system (partly to get the military aboard the program) in the same package, and selling it as the solution to all government spaceflight needs. Imagine instead this stack: a heavy-boost stage of fuel tank +SMEs (make the engine pod recoverable); and an *upper stage * – out of the way of falling ice and foam – divided into two modules: a varying size cargo module for heavy object lifting, and in front the spaceplane with crew compartment and a *smaller * cargo bay for lighter loads, returnable experiment modules, and the manipulator arm. The smaller crew vehicle may have been easier to make more sturdy (or make more of).
The sad part is, there could have been development of a good enough shuttle successor all alongexcept for the inconvenient detail that it would have meant announcing the eventual retirement of the existing shuttles and NASA knew (a) it would look like an admission they had overreached with the STS and (b)they would still not be able to “sell” to the Appropriations Committee going with something just “good enough”.
For some reason, NASA’s institutional culture was that there could be only one major manned spaceflight system going at the same time, that nothing flying under NASA aegis may compete or cast shadows upon anything else programmed under NASA, and that any major change in MSF systems must involve a leap of several orders of magnitude (many worthy incremental-Gemini-derivative proposals were nixed in the 60s because that would be competing w. Apollo and you Did Not Do That).