Fine, I concede the point to you, however, I amend my question:

What actual proof does this dude have that performance and memory of pocket-sized computers have started to slow down?

And how does arguing this point have anything to do with his overall "thesis"?

Forest for the trees, man, forest for the trees...