Having made, and modified RPGs for over 20 years it is always a struggle to keep flexibility while discouraging people trying to take advantage of it. Here is some general thoughts on the subject.
One thing that can be done is make it a matter of diminishing returns or even more so (this is already done to some extent). Go too far up or down is “inefficient” and VERY costly for both the advantages and disadvantages. Also the amounts needed to raise a bonus are figured by the TOTAL bonus a civilization ends up with and not just the amount of bonus given at that point. (this will stop the double dipping problem)
For example, there does not have to be a one to one relation to the extra points you get from disadvantages so picking them could be made as always a last resort “less efficient then just using the points given” to getting the next notch up on a wanted bonus. So in the end picking allot of small disadvantages or even one very big one is “wasting” some points making the process less inviting and mainly for people wanting to really customize and mot minding a slight disadvantage (they might gain that back in play style, or may not, but massive deviations always end up with less total bonuses)
Also if your political group already gives you a bonus in a ability then you are already at that level when you start picking additional bonuses. So a +20 in tech from your political group means you already at that “level’ of bonus and it will take 4 more to add 10 more points right off that bat which is the max. (note that after writing this and not playing around with the bonuses too much I realized that I really have not tested if this was the case or not . .

so, please bear with me )
Also don’t allow beginning bonuses pick totals to be raised or lowered past an unbalancing or ridiculous amount. (ie. The nearly useless scout with a scanner range of 1 or a crazy one of 40)