What you'll earn by having this probably-prototype put into production is tons of valuable data about what works in production, what to improve, what components fails first, how that single-airframe stuff can be made cheaper and more reliable etc. Based on this data nobody else will have, engineers will be able to design weapon systems (think even ML-asysted design based of v2.0 based on telemetric data gathered over decades etc.) nobody else will be able to compete with even if F-35 is a flop. I imagine this is why people fret about the Turkey thing, all this is about the data (that you wouldn't want Russian systems to get a copy of) not about the flying brick.
Overall, putting boring reliable tech in production does not generate much useful data. Putting unreliable prototypes does. And it's all about techno-military supremacy long-term game here, not losing or winning one little war in a country you don't care about.
Now about the pilots and soldiers... yeah, you're doing R&D in "production" (where "production" == "warfare" here), (the wrong) people will be killed, there will be a blood price paid for all this technological advancement, but in the end you'll benefit from it.
...only thing I find slightly horrifying is that this mindset (of "screw the 'client', let'd do some R&D in production to get unique insights and special competitive advantage from it" - that has already sipped into commercial hardware and software for a long time btw), creates an incentive to "gather more and better data" and when this translate to "wage more war" it doesn't bode well...