Shouldn't build up any more heat, and maybe a bit less, than grinding on a standard flat platen, with this config.
You still will have to dip the blade regularly if grinding hard.
I've got a water cooled flat platen (Nathan's with a chiller) and while it really helps prevent heat build up when grinding really thin stuff, most of the heat in the blade isn't coming from the platen or the belt itself, it's being produced by the grinding process itself, i.e. abrasive grains shearing metal from the blade, no matter how you slice it (pun intended) you're going to get heat generated from any type of machining process, you can remove it quickly with coolant as an example, but there's no way even a liquid nitrogen cooled platen would keep heat from being a concern when finish grinding, because the belt isn't conductive enough to carry away the heat quickly enough.
On top of that, in this application, you've got rubber (an insulator) between you and the platen, adding another layer of isolation between where the heat is being generated (on the grinding surface of the work), and the potential heat sink of the platen.
The cooling of the platen is just to keep it from getting hot enough to melt the rubber and pop belts I'd imagine, no doubt it's effective, but necessary because of friction to keep this thing from wearing out quickly.
Demos of this type are often done with non hardened blanks, that are never intended to be finished. Even hardened, if your working with a thick blank, for the sake of demos you very well may not take it down to what most of us consider to be acceptable thickness at the edge before sharpening. I wasn't there, so I can't speak to that, but I have zero doubt that you'll still have to dip a 1/8" thick or thinner blade that's hard, regularly when finish grinding.