Are you thinking about making one vooodooou? Or just making an observation?
Belt drives are utilized in many apps and no sense reinventing the wheel; they are, of course, adapting already existing parts. Trick is, this stuff, if not in the same factories, is being manufactured in the same regions of China; they can cheaply have the specs of their product altered to suit their needs. This alteration may or may not be apparent upon visual inspection (just 'cause they look the same....), furthermore, they may or may not be alterable to the end user in the way that they were to Grubee once mfg. is complete.
I only say this because, yes, w/ off the shelf parts, it is definately possible to make a working transmission. Firebelly and EZ do, Dax and Gebe I assume do. But it's the non off the shelf parts that all of these contain that take the effort to engineer, and therefore, make it so that if the end user manufactures one it is entirely more expensive (or time consuming) than just buying one.
I fully encourage everyone who wants to, to do so. The experience gained is likely to be immense, even when engineering/ reengineering just a few parts. You may have great ideas to offer! I guess just be realistic about your ability and resources to do so.