I think you need to keep in mind that computers take a LOT longer to build than research facilities. (It's a factor of 6 difference, if I recall correctly). So while you might ask when (1.1x > x + 500) and say that level 1 computers are better to build than level 1 research facilities when x > 5000 (in other words, when you have 10 level 1 research facilities), that's not usually the right question if computers take six times as long to build--the right question is when is (1.1x > x + (500*6)), which comes out to SIXTY level 1 research facilities being the switch-over point. If you have level 1 research facilities and level 3 computers (which is normal for me), you should have twenty research labs before switching to computers.
However, the math changes once you run out of room. Then the "factor of 6" goes away (at least mostly).
So what I often do is build research facilities until I hit the breakpoint (usually 20 for me, as I say above), start building computers until I run out of room, and then start SCRAPPING research facilities and replacing them with computers.
How abusive is THAT?
Saben: it seems to me that the morale, race and culture modifiers cancel out. To modify the calculation above, if I ask when (1.3x > x + 500*6*1.2*1.2), I get 14400 instead of the 10000 I got before, but now I need to divide that 14400 by (500*1.2*1.2) to turn it from a number of research points into a number of research facilities, and that gets me the exact same value of 20 I got before.
And my classmates said that algebra was useless....
