Awwwards is not at all representative of the web at large. The set of problems of most websites are almost entirely disjoint from the set of problems on Awwwards sites.
I would also say, in response to one heading in this article—the numbers do lie. The studies it alludes to are somewhere between old and ancient, and being taken significantly out of context and applied far beyond their actual studied scope. The Amazon figure especially is transparently irrelevant in the context of this article.
Yes, things are stupidly bad, but unfortunately this article is shallowly bad too.