Well, I voted yes mainly because of the required (but really boring) micromanaging in the early/mid game. At this stage, I could re-assign citizens to different working tiles pretty much often to optimize shields or food or a very certain mixture of both - and that's just to try to reach these discrete factors (or "production/growth quantums") - eventually turnwise. Mostly it's possible to come at least that close to these discrete quantums that waste is almost zero, there often actually are alternative tiles that could possibly be worked or could be switched/changed due to overlapping for a decent optimization. Maybe I could fill the food storage of a city with 4 fpt for two turns, then reduce to 2 or 3 fpt with a slight (but "quantum"-relevant) spt boost from that mined hill etc.
It does make a difference if your city pumps out 1 warrior/turn (at least sometimes due to short-term shield optimization) or 1warrior/2 turns, or grows in 2 instead of 3 turns in the early game. If you then get "the big picture" (all of your cities), those one turn earlier born citizens might get you that piece of commerce that would be needed to bribe an ally for a MA in time to help to fight your much stronger enemy, for example. Or boost your research rate. Have one extra defender earlier against those nasty barbs in case of shield optimization. Whatever.
So micromanaging does indeed help and I have a very hard time not using it on deity (perhaps I just need exhaustive mm here because I screw up the game elsewhere...).
I could get away on emperor if I just concentrate mm on some selective cities.
Well, I like some "core" mm, but not hard-core mm. So a shield/food cascade might indeed help to reduce mm: food quantum overstep will be stored as bonus towards the next growth process and shield quantum overstep will prompt me for a new production order - so I eventually zoom to the city anyway.
If the result of severe mm was almost no waste, then cascading would just help and doesn't add advantages to a certain civ only.
Later in a game, I would't care for mm (i.e. cascading) in the core that much. Once a city has reached 12, alternative working tiles have become rare anyway. I normally don't need to produce settlers/workers often then (pop status doesn't fluctuate), so I just minimize food production/replace some irrigation with a mine .
Then, if 3 tanks arrive in 8 turns or in 9 turns is not the biggest deal IMHO. But surely a cascade system would still help for newly founded/captured cities.
With regards to all discoveries of corruption and waste - note that the corruption formula just calculates the "red" shields - but the *effective* waste may be even higher due to the quantum system. A city with a raw 10spt and 10% waste (at the best mm result possible) gets a "quantum penalty" for a 10-shields worth build, and has effectively the same productivity as a 10spt city w/ 50% ([yeah, that's the ticket: /] a well fitting propaganda example

).
And that's just shields and not food, the killer resource in the early game.
As for the weaker vs stronger civ: I could imagine that the weaker civ starts in a location that would not feature the desired 5 fpt for a settler factory, but actually has more than 2 fpt. Why should the weaker civ get a quantum penalty on top of
not having an ideal food production (while the settler factory city would not benifit from food cascading at the same time)?
Granted, a cascading limit must be implemented. So no 1-turn build-ups accumulate to a wonder over the times.
And for the beakers - theoretically, I could calculate the required number and optimize the science slider (plus scientists). No fun to do that - and I might even re-adjust things due to drop in price. How would anyone like to have more than one scientist (in case of running science at 0%) when they have a big waste potential (i.e. hard to control them), too?
Well, finally, if the poll results in a draw (or at least isn't highly biassed towards either option), a check box at game start ("like culturally linked..." and stuff) could really come handy. Guess what? It should be automatically checked if using default rules...
