There has been a ham fisted approach to this process that seems to be based around the idea that UI operators should market their own projects, yes - but they haven't really done that. It makes far more sense to have a centralised educational system or to at least make tools available that anyone who runs a Hive UI can use to speed up their own educational system. Most of the features on Hive are shared by all UIs, so it would be pretty straightforward to create educational videos about them.
This situation is similar to the way that most Linux OS projects share the same underlying modules and technology but then customise them and add on their own front ends. While it's true that the various front end projects (Ubuntu, Fedora etc.) do produce their own educational and promotional materials, it's also true that there is a vast array of educational material that is available from other sources, which applies to most Linux based OSs.
If the DHF is going to fund numerous, competing front-ends then it has already committed to supporting these projects and it makes no sense to avoid using funds to create educational material about them - perhaps in a separate, dedicated project.
There are many quality video producers in most regions of the world, the problem is not so much finding them, but in linking them up with people who understand Hive deeply enough to create valuable content, who are great communicators and who have the time / resources / funding to do so. Typically, the whales seem to expect community members to do all the work for next to no pay - this is mostly unrealistic.
Video isn't a bad idea. Video has a short lifespan and training materials can become obsolete rather quickly.
We have countless writers here.
I think the cheapest, most efficient approach would be to use these materials to train a chatbot. Then tie that same AI into everything Hive has to offer.