Plenty of non-taxonomic challenges have been run in the past and gone quite smoothly. All that matters is specific rules are set at the beginning to dictate what does or does not count. For example in this year's nocturnal animals challenge globally, there is a long list of groups of animals that qualify, but the big difference is this group includes some mammals, some birds, some reptiles, and some fish. Something like this would be a good way of doing things, as it ensures everyone a somewhat level playing field between taxa they are particularly interested in and taxa they are not interested in, as well as taxa they see frequently and see infrequently. A lot of those above also do have regularly found definitions that can be used. For example, using a geographic or biome theme, there are readily available boundaries that determine what would or wouldn't count.Most of these require a lot of work on Sealy's part. Taxonomics have easy cut-off lines and anyone can quickly look up if a species counts or not. With most of these, you get in to does xyz count, what if they occasionally eat something else, what about birds that fly over the area, what counts as a desert, what if they used to live there but are locally extinct... I'm not a fan of the discovery criteria, either, since that always means discovered by non-natives, essentially.
My big issue with taxonomic challenges is that they almost always advantage and disadvantage some region of the country over another. Since plenty of zoos have certain taxa they specialize in, or exhibit a number of species from, it's almost inevitable that people in some part of the country will do better just due to what's nearby. For example, if it was ungulates, those of us in the Northeast would be at a big disadvantage, especially compared to those in places like Texas that have significantly more ungulates represented. By providing a wider and more diverse array of what counts makes it less likely for one region to pose a big advantage over another region.
For an example, one of the one I suggested was "Semi-Aquatic Animals". If this theme was selected, it may not account for every possible semi-aquatic animal, but a list could be posted at the beginning of the different taxonomic groups within that can count. It could read something like:
Hippopotamuses, Otters, Beavers, Polar Bears, Pinnipeds, Penguins, Charadriiformes, Anseriformes, Crocodilians, Chelonians (excl. tortoises), Amphibians, and Lungfishes. (there's obviously some more that could be included as well, this was just a quick list of what was on top of my head). If this challenge was chosen, most Zoochatters would be able to find *something* they see regularly included, along with at least some sort of taxa they are interested in. Unfortunately with strict taxonomic challenges (i.e. amphibians, parrots, etc.), I find those challenges to be overall less interesting and can be discouraging from the start if it's a taxonomic group that you either don't see regularly, or aren't interested/knowledgeable in. I ran into this when it was the amphibians challenge, since a few of the zoos I frequent either don't exhibit any amphibians or had the same few common poison dart frogs and that's it. I ended up doing alright in it due to some trips that year (an unusual year that I get to a few major American zoos), but had I not had the means and luck of travel, I would've done extremely poorly and ended with a single digit total. With a taxonomic grouping, that's bound to happen to someone, just on account of where in the country they live.