I think it’s pushing it to say that “if they wanted to, they’d destroy us and be rid of us, and we couldn’t stop it.” In a one-on-one cage fight, unarmed humans lose to most Pokémon, sure, but it’s not like there aren’t plenty of animals in the real world that can tear us limb from limb. I don’t think Pokémon would have the destructive mentality or the organisational capacity to prosecute a war against modern humanity with all its technology, even if they thought it would be beneficial in the long term – and that becomes doubly true when you remember that we’re talking about hundreds of species with zero allegiance to one another. It’s difficult to argue, for instance, that your naturally gentle and placid Pidgey could easily kill you and return to the wild if it was sick of you; it’s equally difficult to claim that a rebellious Tyranitar, after blowing up its own trainer, would stop to free that same Pidgey on its way out the door.
On the other hand, it seems entirely plausible to me that early human societies in the Pokémon world might have thought that this was something they needed to worry about. The first Pokémon trainers were probably seeking protection from very powerful and unpredictable wild Pokémon – and their partners may well have been looking for the safety in numbers that came from the humans’ talent for organisation. That foundation doesn’t necessarily apply in the same way across time and space though.