I'm having the same thought.
Like first of all, we can already generate electricity from evaporation. They're called steam turbines and they're very old technology.
These work at STP, but so do steam turbines if you use a mirror dish to concentrate sunlight.
Secondly, this is just solar power. It's just photon -> heat -> movement -> electricity instead of directly photon -> electricity. My naive take is that it will never be as efficient as solar due to the inherent losses in converting energy.
Lastly, even if we did want to something like this, these birds seem like the worst way to do it.
Couldn't we tap into the same energy source by putting something lightweight and porous like aerogel or cheesecloth onto an "elevator" of sorts and moving it up into the clouds and back down? I.e. it starts off dry at ground level, weighing 10g or whatever. We hoist it up to the clouds where it soaks up water until it weighs 2kg or whatever, then we lower it and harvest the potential energy. Then we wait for the sun to dry it out and we do it again.
It weighs 10g on the way up and something much more on the way down, which I think should net us energy.
To be clear, I still think actual solar power makes more sense, but I don't think grid-scale deployment of this would involve literal drinking birds.