I am tempted to try this on some unfinished projects to try to suss out their flaws that I only subconsciously am aware of.

In a similar vein, I start projects by writing a README. It forces me to see with the user's eyes, which makes the difference obvious between what I think is a good interface and what is actually a good interface. It has worked well. I recommend it!

There's a big flaw in your method: users never read the README.

Github displaying readme.md by default has probably boosted viewings of readmes by a million percent.

Whether reading of readmes has gone up is still an unsettled question

what does .md stand for?

markdown

I always at least scroll through it and enjoy it when someone took the time to write a great one.

When writing a work of fiction, the aim is to find truth not in the creative process, but in the un-dogged pursuit of verisimilitude - in the attempt to build reasonably causal events in response to the creative flights of fancy. Theoretically, it should be possible to achieve this in the frontal lobe, but putting pen to paper has always been a more effective approach at striking the perfect balance.

I'd say the author here decided to apply the same muscles that balance creativity and verisimilitude to his own review. Imagining the response to an unknown, unpublished novel is obviously a work of creativity, which he is also able to balance out because he actually knows the novel, knows its strengths and weaknesses.

In this way, he was able to find a simple, comforting truth: he was either about to become a novelist, or he wasn't.

Writing a piece of pseudo-fiction like this review seems like a great answer to allay the fears of failure.

Writers can be so critical of themselves, especially good ones, I think. Barnes' solution of writing the critique he fears is pretty good. They're only novels, after all, but when you write one, it seems so damned important.

Because it's a neat marketing trick.

I didn't catch where he described it as marketing trick, but I also didn't manage to find anywhere in the article where he actually answers the question in the title

With HN as your audience, who needs professional critics? :)

It can be kind of a relief to freely acknowledge the reasons your thing might suck, one way or another. Ideas can somehow be less scary written down than when just hanging around as possibilities in your head. If you think your bad self-review is the only reasonable way to look at things, then maybe you have a real problem, but hopefully your reaction is to put it in perspective somehow--to consider the good things balancing the bad, or why the objections aren't valid or aren't fatal, or at least that the world keeps spinning and you are not your work. And then when life or other people's reactions come around, your reaction is a bit less surprised and maybe more grounded.