I'm newly signed up on homebrew.stackexchange so I'm limited in where I can contribute. This is not so much an answer, but an elaboration of the question.
I was thinking it would be useful to understand an ideal water, sugar, yeast system. For wine making, one might assume 80% sugar conversion during primary ferment. You'd need to estimate how much yeast is produced, and how much precipitates out of solution. The discarded lees will have a certain minimum hydration level, which you'd need to input empirically.
Perhaps it is hard to estimate total yeast production. What is the rate limiting parameter? Is it yeast population density, or one of the yeast grown reagents (oxygen/nitrogen/trace nutrients)?
Skimming Yeasts & Primary Fermentation, this text indicates that cultured yeast fermentation usually starts around 5 billion yeast cells per litre and hints that available oxygen is one of the primary growth limiting factors (for fermentations that don't get stuck). It goes on to say that total yeast populations for wild ferments tend to be lower than for cultured ferments, so I'm guessing that lees production would be less as a result.
A kit wine maker I'm presenting using allows for an addition of 920 ml of water to a fermentation that begins at 23 litres, to make it up to the desired end volume. I would guess the 920 ml covers various process losses, such as CO2 release, lees production, and evaporation. These kits are engineered to known initial parameters and use extremely predictable yeasts. They have less control over brew temperature and hose jockey slurpage ratios (the recommendation is to slurp from primary just about everything that will go up the tube).
I've consistently found with these kits that I'm about a litre short from 2" below the bung after regulation top-up, even with aggressive slurping. My two recent vintage Italian carboys measure 23.9 kilograms of room temperature tap water (+/- 20 grams). The extra 900 ml of head space is really driving me crazy. I've seen it written elsewhere that recent vintage Italian carboys average closer to 6.25 US gallons (my measurements work out to 6.3 gallons to my desired fill level for secondary ferment).
Anyway, I wanted to look at the fermentation math as part of my dialog with my kit vendor and I found this post. It does seem rather empirical for less controlled scenarios.
(I'd be grateful to anyone who votes me up to the normal minimal level of useful participation. Less crippled I can earn my keep.)