We sometimes use 2 Zambonis at once - but it is only rarely done with the most experienced drivers, when we are in a special hurry.
Fast moving robotic ice surfacers could be very dangerous if someone sneaks in or otherwise enters the ice with them. I am also certiain that smart kids will hack the guidance systems to create chaos. But let's ignore safety issues.
I also wonder if something moving that fast can do a good job. Plus, no matter how you shave off the top surface and lay down water, you really need a few minutes for the water to freeze into a hard, smooth, durable surface. At my facility, even one Zamboni can already do the ice, followed by shoveling the snow while the ice freezes and hardens, in about 8 minutes, using a reasonably experienced driver on the NHL and Olympic surfaces. Small surfaces sometimes take longer, because of Zamboni maneuvering problems. Another facility takes about 30 minutes, and likes to claim that leads to smoother ice. (They can afford to. They are subsidized by a local government, whereas we are commercial.) Trying to get everything done in one minute is asking for imperfection.
Using GPS for this is silly. You could have better accuracy using local fixed transmitters - and perhaps they could be cheap acoustic transmitters instead of radio transmitters.
Maybe it would make more sense to move a bunch of little automated resurfacers back and forth across the ice, smoothing the entire surface in one pass. The low part of a wall could roll out of place. They could be laser or otherwise guided to drive straight across, towing the cable, then be pulled back by the cables and winches, and the wall rolls back in place. With a little work you could automate moving the goal nets around too.
The thing is, most ice rinks already operate very close to or below the economic break-even point. There isn't room in most rink budgets for a lot of fancy equipment and experimentation. And the fancier and faster you get, the more maintenance becomes an issue.