> The issue is that most of the joint programs fail because they start with such a broad difference in requirements the program can't work.
Agreeing on requirements is the most important compromise. You can always come up with different requirements that make collaboration impossible. Sometimes these differing requirements are fundamental - e.g. the E-2 and E-3 (E-7 in the future), while performing similar roles, are necessarily very different platforms - that a joint program would never be considered. Sometimes these requirements are not so fundamental - e.g. the many years of single-engine versus twin-engine disagreement - that a joint program is possible, albeit challenging.
While the JSF program has had plenty of issues, the fact that it has been successful in delivering the envisioned aircraft is a testament to (a) technology improving to the point that many once-dealbreaker requirements (e.g. single-engine) could be swept aside, (b) the rise in costs making it clear that completely separate development programs were impractical, and (c) explicitly providing requirements flexibility via the "family-of-aircraft" approach (no way the F-35C would have been acceptable financially to anyone other than the USN, no way the F-35A would have been suitable for carrier use).
> ...lack of supplier diversity will be a forcing function in the future...
The consolidation of the 90s and 2000s is directly the consequence of lower budgets. Consider the fighter/strike fleets of the 1980s:
- USAF: A-7 (LTV), A-10 (FR), F-4 (MD), F-15 (MD), F-16 (GD), F-111 (GD), F-117 (Lockheed)
- USN: A-6 (Grumman), A-7 (LTV), F-4 (MD), F-14 (Grumman), F/A-18 (MD)
versus today:
- USAF: A-10 (none), F-15 (Boeing), F-15E (Boeing), F-16 (LM), F-22 (LM), F-35A (LM)
- USN: F/A-18E/F (Boeing), F-35C (LM)
there's just not enough business to support much more than the Boeing/LM duopoly keeping both production/support running and enough capability to be competitive for the NGAD programs.