This doesn't mean I think it's unsalvageable - but based on the 2017 Unearthed Arcana the disciplines need shredding and we need to not have e.g. 27 power points at 5th level with a limit of 5 per spell and three disciplines known, therefore a potential of 15 different special abilities not counting talents or subclasses. You've at least as much fiddly stuff to juggle as a wizard does - probably more given that (a) you get the powers in great big gulps rather than a couple of spells at a time and (b) not only do you have these powers but for many of them you need to decide how many power points to spend
Sure, I just don't even think it's hard to salvage it. The basic approach is right, just cut it down - though not as severely as you're suggesting because you're overly concerned with this tiny minority who get analysis paralysis at all, we're already down to like 5% of players, and of those, many can get it on everything.
I dunno if you saw Thomas Shey's posts about analysis paralysis in his group - they have players who regularly get analysis paralysis on like, choosing a feat, that last for two weeks.
Not every class is for everyone. If you must have huge flexibility, Warlocks aren't for you, and nor are most Fighters. If you suffer from huge analysis paralysis, probably forget Wizard, Cleric, Druid and Psion. And so on.
It adds a ton because it's a completely different approach and one that actually jives with approaches from fantasy fiction, unlike all other D&D casters (very much including Sorcerers).
Being massively better than Shadowrun is not being perfect, and keeping the classes down is one of the smart moves 5e has made.
I don't agree re: smart and your argument neither makes obvious sense logically, nor is supported by evidence, and I have a huge amount of anecdotal events that run directly against your claims. I would got as far as to say that in an exception-based system, claiming that more classes pushes players away, when all the previous editions of D&D and every other exception-based class-based game I can think of has tons and tons of classes and does great.
What makes it particularly unconvincing to me is that it's the most casual-ass players who pick the weirdest classes and subclasses, utterly reliably. I hear this ridiculous claim like "new players want and need to play stuff like Champion Fighter". It's complete bollocks. New players are bored stiff by stuff like Champion. New players frequently pick complex and involved classes and engage with them really strongly - like Bard - Bard is not a simple class, it has a lot of moving parts and a lot to consider. Wizards too.
Pure anecdote (but that seems to match with what you've got) but I've seen more players driven away from D&D by it not having a good class to support their concept, or because they hated the mechanics on D&D casters, than because the classes "overwhelmed" them. 2 vs 0 so low numbers but w/e.
I have seen people driven away from D&D by complexity. Absolutely never in the class department though - always in the actual rules.
Do they? They've been adding psionics - but the psychic warrior is an explicit fighter subclass.
Yes. They've repeatedly said they understand people want it. That's why they tried adding it early, like in every other edition since 2E. But under Mearls they used the utterly fatuous 70% test which would never have allowed any full caster to go live as it actually is in 5E. Warlocks, a simple class, would definitely have been stopped - the objections to Eldritch Blast and esp. the enhancements to it would have just been endless. Wizards even would never have made it, and especially not their subclasses. God help Druids or Bards, they'd be right out. Clerics would have been lucky if they did and many Domains wouldn't have made it.
And I've pointed out why the mystic (as of 2017) isn't even close to being there. It started on the right path and then massively overcomplicated things to the point it reads to me like Analysis Paralysis: the class. Which is the opposite of what I think we both want.
You're overstating and overdramatizing. It could stand to be cut down a bit (there are too many abilities that do essentially the same thing), but again it's exception-based, you only need to know what you need to know. And if you are a person with a problem with analysis paralysis as bad as you're describing, there are bunch of 5E classes that will be an issue.
Could it be more simple? Yes. The opposite? Absolutely not. It's most of the way there. Even you say it starts on the right path. I don't see how that can be compatible with it being "the opposite", rather than "not good enough". The opposite of what I want is Wizard or Sorcerer just with a psionic subclass, being sold as a Psion. That's "Kill me now" stuff.
I was not joking when I called it analysis paralysis: the class and it deserves to go into a museum of traps not to fall in when designing a class.
This is just overcooked nonsense and you literally haven't made a single rational argument in favour of it, just presented the same opinions about how it's "bad" repeatedly.
they've always struck me as being wizards with different hats
Which demonstrates a profound lack of understanding of the issue. It's like thinking Paladins are just "Fighters who are a bit religious". Or Rangers are "Fighters who crap in the woods", except even more extreme. Your "argument" is just a lot of extremely strong opinions about something that doesn't even really impact you, because you'd never even play it.