Well, useful is always in the eye of the beholder. There is no such thing as an absolute truth, after all. And pretending there is, and it's even attainable, is intellectually dishonest.
Sure, an essay could be a formal piece that approaches an almost "mathematical" approach. After all, an essay a first and foremost an argument presented by the author. Even a flawed argument is still an argument. And a flawed essay is still an essay.
The fallacy here is being implicitly reductionist. If your premise states "an essay should be useful" then you're basically reducing the definition of what an essay is to a formal argument based on logic and falsifiable facts, and rejecting any other text as "not an essay" or, worse, "not useful" - whatever that might mean - or, worse, "nonsenses" or "a dumb thing to say".
A quick glance on Wikipedia dispenses such reductionism rather swiftly:
https://en.wikipedia.org/wiki/Essay
Not-withstanding, I think PG's essay does contain some excellent personal advice on writing style and technique itself. No more, no less. His sin is confounding form and function. The former always follows the latter, never the inverse.