If you want more granularity, the easiest way is to increase science production and cost.
Example : multiply by 10 all science production and tech cost
(and divide by 10 the importance of science when comparing science yields to other yields)
Remark : 10 is probably too much. But 2 or 3 would probably help.
You are really suggesting real numbers: 1.5 , 1.2 and so.
I think the problem is not there. It is what is tied to the bonus. When it is tied to population, it rewards growth, thus food. When it is tied to city connections it rewards wide empires (so good to have it locked inside Progress). When it is tied to border expansion it rewards a large empire only when it is expanding, and more if it has some culture, but it stops before late game. If it scales with era, it rewards fast teching. If it is tied to resources, then it is a lottery. If it isn't tied to anything, then it is a flat bonus, very rewarding early and negligible later, no matter the granularity.
So, how can we scale science bonus in a way that doesn't unbalance everything? We can't. Every system has its own balance.
Linking to the pop size was a bad idea, but everything scaled to this. I don't know yet if I like the yield per event, I haven't tried it enough, but understand the concerns.
A couple of ideas of what to do with those science buildings:
*Make the library/university be required for 'turn production into science' build order, and let it be bigger (75% hammers into beakers). Optionally, create a new build order that turns gold into science (actually it is purchasing beakers), available once library/university is built.
*Make the university/public school to increase +50% beaker production. That's to be added to the 'hammers/gold into beakers'.
*Building Council/Library allows trade routes to "move" beakers, as they do with hammers and food.
So a city focusing in science cannot build anything while researching, or is going to be short of gold, or both.