There are other animal models, some of which even seem to develop dementia-like symptoms spontaneously.

Mice have absolutely dominated research because they're relatively cheap, lots of powerful genetic tools are available, and the PR is more tractable. However, that doesn't mean they're the right choice for every experiment.

This argument is like the infamous drunk who was arguing that he should continue looking for his keys under the street light, because the place where he actually dropped them had no light so he couldn't see.

At this point it is a waste of time and money to produce more drugs that are guaranteed to fail in humans. Making it harder to waste time and money in this way is a feature, not a bug.

After we have better basic science about what likely causes Alzheimer's, we can go back to the question of how to find drugs that might help. Trying to do it in the other order is putting the cart before the horse.