Good and interesting points...
One basic issue is that STOVL is difficult, which is why it has taken 45 years to get to a practical supersonic STOVL aircraft. The bring-back load is still a small fraction of the total landing mass, so that any problem or deficit can wipe it out rather quickly.
The cost and lack of flexibility involved in CV is real. That could be fixed with automatic landing, but that hasn't been demoed yet (people on real programs tell me it's possible), as could the lifetime-eating nature of CV ops. But as it stands I can see where STOVL is the only solution for a joint force as the RAF and RN plan.
But - Enfield Schmenfield - show me where troubled programs insulated from outsiders have ever put themselves right. Unquestionably it is up to the qualified to act and to make decisions. However, military operational experts, procurement insiders and engineers brought us the Boulton-Paul Defiant, the K-Class submarine and the Nimrod AEW.3. (And the insiders behind the last were mightily P
ed at the World In Action crew who blew the whistle.)
And somebody here brought up the delays to Typhoon... which were publicized and addressed in the mid-1990s, by journos looking down the barrels of the Baron's legal goons, including the legendary Carter-Ruck. And what had suggested that there were problems? If you look at the better contemporary reports, it was that, a year or more after first flight, only a few dozen hours of testing had been carried out.
And of course the insiders said nothing was wrong, that the journos didn't have the qualifications to talk about the subject, and (above all) God forbid that they had found a black hole in the envelope that the jet couldn't get out of. (Cough) ALSR (cough).