Across Eurasia you had about 2000 years where people were riding around on chariots with strong spoked wheels, but farms didn't have even primitive wheelbarrows to move rocks off their fields. Moving rocks off fields is back breaking work and even a primitive improvised wheel barrow gives you an order of magnitude advantage. The prerequisite technology, conditions, demand, etc were there, but it took thousands of years for somebody to have the idea. Therefore, I think it's safe to conclude that very few people who've never seen or heard of a wheelbarrow would ever have the idea to invent one, even if they were already familiar with wheels in other contexts. The idea itself is the rare thing, not the prerequisite technological/social conditions. The idea can go unhad for thousands of years after all the other pieces were in place.
> You aren't even addressing Diamond's argument.
I think Jared Diamond could be correct, but I would still say that his argument is lame because he failed to establish a high likelihood of prompt invention once prerequisite conditions are met before concluding what those supposedly missing prerequisites were. I'm not addressing his argument because he failed to convince me that his argument addresses a real question in the first place. Supposing hypothetically that ancient Mexico had pack animals, it still might have been thousands of years before pack animals and wheels were put together.
> Going from [rollers] to the wheel doesn't seem like a huge leap of insight.
I really think it was. There's a big conceptual gap between rolling something on logs, and making the rollers captive through the use of an axle. The roller goes from being a tool used to move a thing to being an integral part of the thing. Wheels may have been invented more than twice, but there's not much evidence for that. And in any case, it certainly did take a very long time for some seemingly simple applications of wheels to be considered after wheels were popularized for more sophisticated and demanding purposes.