I wouldn't put too much store by these statements, without much fuller context. A range of origins & processes at broadly similar roast levels will give you a 3% wide range of EY for even a very consistent pour over method. Many quotes along the lines of "I got this % EY", don't seem to be average EYs with a reasonable tolerance of extraction.
Over-extraction is a thing, it's the thing where extracting too much from the grounds causes smokey, drying bitterness. If it wasn't a thing, Perger would be telling you how to make 29% or 35%EY drip brews that taste good. No one can seemingly do this. It is easy to make a brew taste bad by extracting less that the maximum, too fine a grind allowing too much silt in the cup, too coarse a grind limiting higher end of tasty, too much turbulence again pushing too much silt in the cup.
As grinds get finer, assuming you can still get percolation without choking or channelling, tasty EY can increase, because you are making the dose more soluble. The 18-22% gold cup range was determined with average grind sizes of 700um & over (and only using the percolation formula). Finer grinds certainly allow tasty percolations at higher EY, than coarse.
The outer layer theory has a few holes, yes coarser grinds extract less, proportionally, but it's not as linear as "to 100um deep".
Dose & water only really affect strength, once you have normalised grind setting. Brewing too fast with a similar extraction? I'm not sure what "too fast" means here?
A grinder that grinds finer & still allows normal flow (doesn't make so many fines that brews channel/clog), can make a higher end of normal extraction (I think) because there are fewer large particles that will have all the tasty material scrubbed from the outer volume, before the rest of the brew hits a high EY. You can test this yourself. Take your typical V60 recipe & grind, make a few brews, now grind more than the dose that you need & sift out the largest 20% & make cups with the same weights & regime, what happens? Sift out 30-40% what happens?
I'm not suggesting that you sift out a large proportion of your grind for brews from now on, just for these experiments. Day to day, find the average EY for your regime/grinder that yields the highest strike rate of good cups & concentrate on enjoying your coffee. If you want to lift that EY to see what happens, grind gradually finer/agitate less, until silt becomes an issue, or more cups than they should start tasting bad.
Do not make conclusions over a small number of brews, nor based on one coffee/origin only. Unless things are really awful & need a quick fix (say I'm trying to dial in an experiment), I wait until I have brewed 10 different coffees before making a significant change. Take account of coffees that are over/under roasted, you can't fix these with brewing.
So, prime take away is, max tasty EY depends on grind fineness. Useable grind fineness is dependent on your grinder. If your grinder works better coarser, aim lower for EY.
I don't have a flat burr grinder, but in terms of preference it's very difficult for me to determine any taste preference between (dialled in) good conical grinders (Niche, Lido, Feld) & cheap conical grinders (Porlex, Rhino, Hario Slim), or even with a blade grinder with a mesh to limit larger particle size.