This ties directly in with the locked vs free compounds stuff that's part of the CPA system -- from the blueprint of the cell we need to figure out how many compounds go into manufacturing a copy. Sadly, we never actually answered that question, I'm only pointing out that the question is the same.
I think organelle upgrade compound cost is a particularly tricky one -- certain upgrades would reduce the compound cost of an organelle that has the same level of function, for example, while others might increase the size of an organelle to increase its effectiveness, thereby increasing compound cost. I think we'd have to set numbers, real numbers, using estimates of the dry-weight breakdowns of various organelles, and fudge the numbers one way or another from there.
We'd also have to improve the compound/process system so we get the whole self-regulating thing going, where ATP levels are regulated by making supply (through glycolysis and respiration) match demand, and where, say, proteins, lipids, and nucleotides/nucleic acids are constantly produced at a certain rate to feed growth, with additional production of those three to feed repairs.
Then, to decide if you are able to reproduce yet, you check if your stockpiles of those 3 compounds are at least double the minima, and if so, you can reproduce, and each daughter cell will have exactly half of everything.
Personally, I'm very partial to a more nuanced system: let the player reproduce even if they have, say, only 1.6x the minimum required protein, but each daughter cell would be starved for resources (ie, organelles would suffer scaled amounts of damage), and maybe they'd have a little penalty to mutation points to discourage gamey speedmutating, if we decide we want to do that. Obviously, we would have to make this clear in the interface -- instead of having a simple progress bar, we'd have a progress bar with a marker sitting somewhere in the middle, to indicate when you can reproduce, and when you can do so healthily.
None of what I'm suggesting is new, honestly, I thought we already decided on all this a few months ago, but maybe you weren't happy with the result so you're opening discussion again
Now, as specific responses to a few of your points:
- threshold size: I assume this means by dry mass. By extension, I assume this would probably measure size as something like min(compound_x_amt/compound_x_min_amt_for_mitosis, ...), which is exactly what I'm suggesting above (barring the nuance about being allowed to split too soon)
- required compounds: also pretty much what I said above, together with threshold size
- ATP at normal level: I like it. Maybe add the additional constraint that a certain amount of glycogen/other-sources-of-energy be available, since mitosis is not cheap.
With all of these requirements, we could probably have the UI be something like a simple progress bar with a little popup display on hover that summarizes requirements and shows which are met/unmet.
Maybe, in addition to all this, we could have each organelle (well, the ones where this makes sense at least) have a 'size' parameter that scales up as protein/lipids/nucleic acids get incorporated in them.