About complexity:
In BG1+2 you chose your class, stats and profiencies at character creation.
From then on you can only spend profiency points and thief skills, nothing else.
The only exception is dual class for humans and this you must plan from the start because of stat requirements and at what level you want to do it.
Well, and spells for sorcerers, that is very importent.
Yes, BG1+2 has trap choices.
A mage with low int, great, you will never learn new spells because you fail to copy it from a scroll.
OK, maxing out the main stats for your class is not really a secret that is hard to find out.
Well, but do you think a mage -> fighter dual class is as good as a fighter -> mage?
In BG3 you will have multi classing.
At every level up you can chose what class to take (there may be stat requirements).
You have to select a subclass for each class and some casters have to select new spells every level up.
Every 4 levels you chose between a stat upgrade or a feat.
So I think DnD 5E has more options than DnD2E but less than DnD 3E or pathfinder.
I have the alpha for the new pathfinder game.
Without any spoilers, when I started character creation I was struck with analysis paralysis.
You have tons of classes and many subclasses for each class, plus tons of feats and lots of unique abilities to select for each class.
After looking through the options for some time I got a headache and quit playing.
Later I started the game again, ignored everything and just re created a character that worked in P:K.
I think there can be too much complexity.
I can understand why nerds love DnD 3E or pathfinder, but I can also understand when many players complain that this is too much for them and they feel lost when they get the task to create a good char.
Having a choice is good, but knowing the pros and cons of hundrets of classes (if you count the subclasses), which ability you get at which level, what works together with something else and so on can be too much sometimes.

Prof. Dr. Dr. Mad S. Tist

World leading expert of artificial stupidity.
Because there are too many people who work on artificial intelligence already