I'll second the p.2) of the previous post by Kiero.
I think it will be great to have more situational research multipliers in addition to research complexes.
Now we have Ancient Constructs that are doing like that, but it might be much more consistent and less micro-heavy, I think, if research model will be more weighted in this manner.
There might be:
1. Instead of fixed amount of RsPs alien component can give while being instantly deconstructed - it might give fixed multiplier for the relevant research fields, while this research is going, with some chance of the researching component being broken in the process every construction cycle. (And that is local multiplier - components have to be in the same colony with research labs to give bonus.)
2. Ground force adaptations. The more population you have in some surface type (grav, atm, temperature, terrain) - the more global research multiplier you might have with this type of GF specs, representing cumulative practical experience your empire have with this surface (LOG function, I think, will be appropriate). It might be local and global both, so it will be very easy for you to research, say, desert low grav infantry, if your home world is low grav desert. In addition, GF construction cost might be lower (spec multiplier halved?), if your GF construction facility is located on the planet with the same surface spec, so it will be much cheaper to build up desert low grav infantry in the desert low grav planet.
3. Some small global bonus when a ship with Science Department and survey officer is stationed at specific bodies (Ancient Constructs and so on), even if there is no colony there.