[akots]

Generally (! but not always !) what matters is the total science output of your empire in beakers per turn and not where your science slider is. With a large empire and at war you can run it at 50% but still make way more bpt than with small empire running at 100%. Also, there is no corruption and with larger empire, you can make more units, have more resources (bigger and healthier cities) and thus do not need too many improvements to make that happen in all cities which saves you hammers towards more useful builds. For example, compare banks with grocers. If you have more health, you don't need grocers and can go directly for banks. But granaries are always needed. You would not also need markets or temples since you can get that happiness out of military police warriors. And so on. Building courthouses and Forbidden Palace or moving Palace to a more central location can greatly reduce maintenance. Or you can play Zulu and get cheap -20% maintenance almost instantly in all cities. Or you can research full throttle to State property and again get rid of distance maintenance completely.

Benefit of GM is usually (! not always !) an illusion since you can get more beakers with GS than running 100% science due to maney made with GM. For example, you run 70% science for most of the game and can run 100% science for a few turns (let's assume 10 turns). So you save 30% * 0.25 (assuming you have libraries everywhere and no markets at all) = 37.5% of your total budget. Let's assume also you are playing on monarch level standard size map and have 7-8 cities in early Renaissance. This makes your total income about 400g and you gain roughly 1500 beakers (gold frm GS mission) this way. Usually, you can get up to 2100 beaker towards Education (which is a reasonable tech to research at this time) with a GS and you get the tech earlier because the gain is instant and does not have to be distributed over 10 turns. This makes GS better in many cases (! but not always !).

Farming GE requires early Metal Casting and early forge and even then it is hard to keep the GE pool clear from any contamination with other GP points. For normal speed this requires running an engineer in a city with Great Wall (1 GE ppt) for about 25 turns for total 3 gpp. And you cannot run any other specialists there and cannot build other wonders. Pyramids are better but on higher difficulty are hard to get and require considerable sacrifice in development which might result in tech retardation and cornering by AI in a closed space. General (! not always !) these hammers are better off to be spent towards military and expansion and libraries. Of course, with an industrious leader and with stone and a few worthless forests nearby, Pyramids can be built early and are well rewarding wonder. But normally this kind of situation is extremely rare. Most of wonders apart from Oracle early on and may be Great Library are relatively worthless down to much later in the game usually (! not always !). Sometimes, it is worth rushing Oxford with GE in a scientce-oriented capital with very low hammers. But one can revolt to slavery and whip that Oxford as an option.

So, screwing expansion and growth for a GE or GM is not worth it generally.