Compared to 3E/5E style multiclassing PF2 offers next to no real customization power. And by that I mean "creating a character stronger in combat than the others".
Examples: 3E allowed you to create a paladin with nigh-impenetrable saves, because the very first level of a prestige class granted an outsized save bonus, for instance. There is zero ways to create "a fighter, but with a better to-hit than anyone else in exchange for a worse AC" for instance. (Instead you're asked to create a Barbarian that's designed that way from the start, which isn't the same thing or even close). Paizo doesn't trust you to decide for yourself which of the three saving throws you want your character to excel at - it's decided by your choice of class at level 1. The end.
After all minmaxing is the art of maximizing your strengths in areas with the greatest impact (most commonly, the ability to turn live monsters into dead ones) at the expense of having weaknesses, preferably in areas with a lower impact than the devs could imagine (through clever tactics, if nothing else).
The archetypal (and probably prejudiced) munchkin example: doing more damage at the expense of being worse at social encounters (moving points from Charisma to Strength). By then never engaging in diplomacy, and instead slaughtering every ugly mug you see*, you have effectively negated your weakness while maximizing the utility of your strength.
My players have found that minmaxing in the build stage is exceedingly difficult in PF2. It does seem like Paizo have failed to make all classes equally strong, and so about the only advice I can give you at this stage is "play barbarians and bards instead of alchemists and wizards". Once you've made that choice at level 1, though, you aren't given any other opportunities to make a difference.
Personally, I actually like that especially items can give you an edge over and beyond the expected math, and this is why I think Paizo went too far. If 3E was broken as naughty word, and 4E was way way too contained, I personally prefer 5E's stance over PF2 in this regard. Put plainly: PF2 contains way too many feats and magic items that remind me of 4E magic items - very bland and mediocre bonuses you don't get unless you're willing to jump through surprisingly byzantine hoops.
*) D&D is positively neanderthalian in this regard - if you look good, you are overwhelmingly likely to actually be good. If you look monstrous, you are quite often literally a monster. And yes, I'm talking about adventures written in 2020, not just 1970. I would say that substituting "detect evil" for "would I naughty word it" doesn't give appreciably worse readings overall... so maybe them munchkins had a point all along?
I find this an odd viewpoint. In 5E, you could create characters with no weaknesses or minimal weaknesses. I did it more than a few times. There was only one or two ways to build a character into something powerful. They were clearly delineated. You took the same thing the same way every time to become powerful. It wasn't like you had a lot of options to become powerful. You took GWM if a martial or Sharpshooter if an archer. You were a sorlock as a caster or nothing. You took a few levels of paladin for smite or six levels for the saves, then did something else like fighter if you felt like it. You didn't have a lot of meaningful choices, you had a few simple limited choices. You wanted to make a two weapon fighter or a knife thrower, don't bother. You want to make a straight wizard that does damage, forget about it, sorlock makes you look weak. And
bless was a requirement. Keep it up at all times in every fight.
Now PF1 definitely allowed far more customization, not necessarily in the core rulebook which is the only book we have for PF2 right now. After nearly 18 years of materials, it allowed an enormous amount of customization. I would say the PF2 core rulebook had more customization than the PF1 core rulebook.
I believe the choices are meaningful in PF2, but instead of allowing a character to become some overpowered, unassailable monster they instead allow you to create a character that does what it does well. I don't consider that meaningless. That seems an odd way to see it.
I created many characters in PF1 and they had no weaknesses. They were unassailable. 90% of the feats in PF1 were pointless as were the archetypes because of how weak they were. There was a way to build a powerful, nearly unassailable character. It would look nearly the same for every single person building that powerful unassailable option. There were very few trade offs. That doesn't seem like variation to me. It was just straight up power gaming. If you made a paladin, you didn't have to do any work to build up yours saves other than raise your charisma. You could go a little farther with save resistance by multi-classing some levels of monk for across the board good saves with evasion, but it wasn't hugely necessary. I built so many power monsters in PF1. There were no trade offs. You took the most powerful options and made a nearly unassailable character the DM hated.
I think you have much greater variation when you can make a game where the player can pick a concept and not feel gimped for doing it. In PF2 you can make a good archer, two-weapon fighter, two-hander fighter, or sword and board guy and not feel like you're the gimp sidekick along for the ride because the other guy built some monster two-hander guy that can't be touched. Then again, who made fighters in PF1 beyond getting weapon spec. That weak will save was a no-go in PF1. Too many will save spells that ended your day.
Comparing PF1 core rulebook versus PF2 core rule book, PF2 core rulebook offers way more customization. Now if you compare PF1 everything compared to the PF2 core rulebook, sure, I see what you're saying. But in all the years of PF1 material they put out, 90% of it wasn't very meaningful. You had about 10% to work with if you wanted to maximize your character, then a ton of options everyone ignored. A new book would come out and maybe one or two feats would end up being used, a handful of spells, and maybe one archetype. To me that's isn't huge variation, that's limited options that you must take or you are weaker than someone taking more powerful options.
Whereas in PF2 you could probably use 50% to 70% or more of the options available to make a class of fairly equal effectiveness to another class in a uniquely different way. And you wouldn't feel like a gimp doing it. I believe that gives players a larger degree of variation than PF1 or 5E in that they can make some interesting concept that you wouldn't normally try because it would be considered an inferior option in PF1 or 5E.
One 5E example is one of my players made a straight paladin. While this other player made a wizard with two levels of paladin and a whip. Here was this paladin whose main ability was to smite to do damage and this wizard-paladin had way more spell slots and a reach weapon to smite with. Not to mention spell versatility with fly. Then this other player in the same group made a warlock-sorcerer-paladin doing smites and eldritch beams. The poor paladin felt pathetic. His entire class over-shadowed by people dipping two levels of paladin.
I still don't get the alchemist hate. The bomber alchemist does a ton of damage and is very helpful. He's been a huge boon to our party.
The only class so far that has been underwhelming is the wizard. The lack of a good one option action and the magic limits just make them pretty lame. They really missed the boat on the wizard. For all the thought they put into balancing, the wizard is terrible. Like they hired some designer that hated wizards and was absolutely intent on making sure they were the D list class this version of PF.