The problem is that certain "ideological" choices are either mutually exclusive (in the real world) with each other, or are significantly difficult to reconcile.
For example, having a strong military where everyone serves mandatory military service for a fixed term is significantly harder to integrate with a belief that individual liberty is much more important than collective need, vs a society which follows fascism. They're obviously not incompatible, but a society that wants both has to work a WHOLE lot harder trying to get them to work together than a society that only wants one of the two.
I'd rather not have to have GC3 micromanage the relationships in some sort of uber-tech-tree web where all sorts of dependencies were laid out.
Rather, I think the rational way to handle this, as well as reflect how a society might change ideology over time is this:
(1) as far as the AI is concerned, you "alignment" is determined by whichever ideology has the highest level choice made first. For example, if you have a level 1 Benevolent choice and a Level 2 Malevolent choice, then the AI considers you a Malevolent society. If you have two Level 3 Benevolents and only one Level 3 Malevolent, BUT you choice the Malevolent Level 3 first, then you're a Malevolent society. In the latter case, you could switch to being a Benevolent society if you chose a Level 4 Benevolent ability.
(2) All abilities in all ideologies are still available. BUT, all costs for abilities outside the ideology you are currently considered to be (as per #1) should be 2x what your main ideology is.
This way, you can mix and match, but there's a cost to trying to integrate all the conflicting abilities together. You can also switch main ideologies under this system, but, once again, there has to be some considerable effort to do so.
The fact is, certain memes and societal beliefs don't mix well. Allowing people to merely hop around the ideology tree without recognizing this issue is a major game-breaker.