The problems you list are (AFAIK) reasonable. Let's see if we can dream up solutions.
Dirt floor. Lay down a plastic barrier and then interlocking plastic panels, not unlike how ice rinks are created on arena floors.
Are blades baked in ovens? I see nothing about that here -
https://www.nap.edu/read/1824/chapter/7
Painting. The outer finish, the gelcoat, is sprayed into the mold as the first step. Then the layup is done on top of the gelcoat. There's some touchup painting that's done after the blade haves are joined, but that's hand work.
Power? Run the transmission lines sooner than normal.
Workers? We now move construction crews from wind farm to wind farm. Housing, breaks to go home to visit families, etc. is already something done.
Management? On site as needed. Skype connections to the main office/engineering.
Water? Supply storage? Minor issues. Water trucks (tankers). A few shipping containers for storage. Bring in what is needed per week (some interval) in a shipping container and haul the empty one away.
If the blade factory is set up close to the wind farm then it becomes affordable to move the blades the last few miles by chopper.
Thing is, we're hitting an onshore limit to turbine size due to the difficulty in moving blades for anything larger than 3 WM. We're installing 8 WM at sea and companies are designing much larger (up to 50 MW) turbines for offshore.
Some good observations. In general I think I'm coming from the direction of a large, established tech giant like GE trying to do this. It is so far outside of their normal way of doing anything that it just won't happen. A small startup, maybe, but then those don't get (multi)billion dollar contracts for a huge wind farm.
Interlocking panels such as an ice rink only work on a solid, high-load capability base (concrete).
While I don't know for certain how blades are cured, that reference is 25 years old and blades have only become larger and more weight/stress critical since then. It is very difficult to get good hot-wet properties out of a matrix resin without an elevated temp cure.
Gelcoat makes sense.
Carting workers around: that is the norm for small outdoor construction crews that are just ("just") digging holes, pouring foundations and bolting together pre-manufactured components, but that's not the norm nor will it be easy for the type of facility envisioned. Not that it can't be done, it just isn't now.
Management by Skype: See small, nimble young company versus staid engineering behemoth above

.
Water, etc. I think I envision a workforce an order of magnitude larger, or more, than you do.
If the whole point of moving the facility (more $$) is to save transportation money (less $$), why does that mean you have more to throw at helicopters? It just means you save the train and truck costs.
But I get the point about significantly larger blades. Obviously they can't get much larger than now and still be manufactured on the east coast (GE, for instance) and be railed shipped to Iowa. They span two train cars now, I really doubt spanning three or four is doable. I'm sure there are a lot of finance people penciling out how much it costs to do what you suggest versus how much more they can charge for larger, more efficient units.
Interesting times.
And discussion. Thanks!