I have a lot of disorganized thoughts, so I will try to have them listed as coherently as possible:
Originally I had thought that the TSMC AZ stuff was way overblown, but there is too much evidence that I can no longer just write everything off anymore. There is clearly some degree of issues (albeit how much and if they are now mostly smoothed over is unclear). What does seem pretty clear at this time is that TSMC didn't do all of their homework before starting this endeavor.
On the topic of hard work, I have a few thoughts. I think a lot of folks had assumed that a job at a chip manufacturer would be closer to a job at Microsoft or AMD, rather than like a job at P&G or DOW. So it doesn't surprise me that there is a sizable number of burnouts when people learn it just isn't for them. The line about fab phones is a good example of what I mean. It isn't just a security thing either, but also my understanding is they pick models to minimize RF disrupting the plasmas of nearby chambers (similar to how newer iPhone radios can stop people's pacemakers). Almost all semi firms also seem to love hiring overqualified PhD people to do BS jobs, and BS folks to be overqualified technicians (albeit the later seems to be less common than the former). If you are a PhD person who wants to do research, doing process engineer/sustaining tasks will definitely be a turn off for those kinds of people. In my opinion this industry wide practice is easily the dumbest part of this whole industry. IMO PhD's should be process chemists/CR researchers/development team integration members only. That way your manufacturing applicant pools open way up, your avg cost per employee goes down, and you end up with less burned out folks who feel lied to. But there is so much bad press that it seems like this goes beyond what you would expect if it was just the above items.
As for the idea that Americans are just not cut out for the discipline or rigor needed to be successful in the semi manufacturing industry I am completely unconvinced. Intel, Ti, Samsung, and GF all have significant US presences and do well for themselves. If they were unsatisfied they would have stopped expanding in the US. Heck my understanding is that when it comes to 14LPP foundry business (Samsung's most succesful foundry node to date) that Samsung Austin was/is the whole network for that node family (and was also where Apple did there iPhone SOCs). And while intel technically hasn't proven anything on the cost effective manufacturing front, I don't think there is any indication that intel America employees work less hard than TSMC employees. Besides the fact that I see plenty of sleepless nights from American process engineers and techs working themselves to the bone in hot bunny suits, I think a good case study is in the area of TD. Intel 3 has if nothing else closed the PPW gap with TSMC vs where they were in 2021. If D1 employees just worked SO MUCH less than TSMC's Fab 12 guys, the only way for intel to close the gap is for their engineers to just be wildly smarter than TSMC's. TSMC's engineers are far far far from dumb (same thing with Samsung's engineers). With that said my gut feeling is that engineers from all three work very hard and are similarly competent, with those little 1% differences/raw execution being what determines how technology competitiveness shakes out. And then there is the elephant in the room; Micron. They compete with Samsung and SK all firms with deeper pockets. And in Samsung's case they are well known to have no respect for their employees and working them to the bone. Yet Micron leads both of these firms. They also navigate wild language differences (ENG, JP, Mandarin, and Malay) in their global manufacturing network. If we have all of these examples in the logic and memory side of semiconductor manufacturing excellence and technology development excellence is the problem more likely to be the American worker or is it just TSMC? My bet is if it has to be one that it is the later being in the wrong not everyone else.
Hiring practice wise in TW TSMC could use their prestige (and lack of non electronics industry jobs) to get away with worse comp/vacation packages. I have even heard that folks like UMC/Micron need to actually have better packages just to overcome that prestige wall (that seems plausible, but I obviously can't personally comment on the veracity of those claims or just how big the gap is). In the USA that is a no-go since there are tons of great options. For example I bet Chevron-Phillips pays more for an equivalent engineer than TSMC. So if I am an American who wants to be a process engineer, I would be stupid to work at TSMC unless they gave better benefits, more vacation time, or if I really want to work in the Semi industry. I think it is possible that TSMC overvalued the prestige of working there and many folks just said "nah I'll continue working at intel, P&G, 3M, ASM, ect".
Final thing that comes to mind reading the linked article is that this feels very reminiscent of TMG in the 2000s/2010s, which gives me bad energy. An environment of obedience is in my opinion not a good place to innovate, as then every idea needs to come from the top down. Now if you have a swarm of grizzled veteran engineers/techs maybe you can get away with that like TSMC has so far done. But even then with fewer eyes on a problem you will likely get sub-optimal ideas, and if you ever have a dilution than the whole system will fall apart. But if you instead have a culture where people are safe to learn and innovate you can over time have more experts and better solutions. The language barrier thing is also kind of surprising (in how bad it apparently is sense because it was a given that there was always going to be some degree of issues from that). Back when I was college senior who was applying all over I thought I remember my interviewers from TSMC explicitly telling me something to the effect that I didn't need to know any Chinese because everything would be in English. If these accounts are true that is just miserable for everyone involved. The trainee will feel like crap since they are having a hard time learning, and their superiors will be pissed at his "slow American transplant that is taking longer to learn this stuff than he did".
As for the TSMC Taiwan folks who are staying in AZ for the long haul I don't really get why folks are that rattled by it. You can't have an efficient fab of just new hires/RCGs. You need an experienced nucleus for these folks to grow from and solve the out of the ordinary issues that you can only solve after having seen it before.
Hope this is at least sort of comprehensible, and that whatever issues TSMC was having have been or will soon be fixed!