>What could explain this sales pipeline, if the F-35 was the boondoggle this article implies it to be?

A lack of actual proven fight-testing.

But it did the most extensive flight test program for anything in history [0]. I worked on this program for years. I do not think a lack of flight testing is the problem. The problems are many, but in short:

1) Lack of competent, forceful oversight from the program office. DOT&E reports about the F-35 program have, for years, given the program office recommendations that it has failed to follow.

2) A prime contractor (Lockheed-Martin) that restricts access to its data. The F-35 program had to sue LM in federal court to get access to the necessary data to make the Joint Simulation Environment (JSE) fully functional. In the end, the case was settled, but only after six years of battle. The report linked in the parent article describes how maintainers are not allowed access to servicing procedures that they have on other aircraft. I have seen this personally in flight test. Even something like a gear swing requires an LM certified Field Services Engineer to conduct.

3) A completely broken software release process. For many years in developmental flight test, we received software builds that were just entirely broken, as in, the jet would not start with that software loaded. The C2D2 process was advertised as fixing this, but really it was just a new name for the same old fly-fix-fly process. The parent report details entire versions that were skipped in the IOT&E process because they were so buggy. The program could have turned JSE into the final stop for new software builds before hitting the fleet, but it chose to pivot entirely into training instead.

I could keep going. A decade working in a program like this gives you a long list of things to talk about. But I'll stop here for now.

[0] https://news.lockheedmartin.com/2018-04-12-F-35-Completes-Mo...