In what world is it a good idea for an engineering director to take a deep interest in technical details of a project? Suppose that the director was right and author was wrong in ~all of their disagreements, but just that the author was simply telling the truth about two things: the director spending two weeks doing design mockups, and intervening to specify the type of a database column. This makes the director somebody who is (a) not doing their actual job, and (b) probably nightmarish to work for no matter whether you are a "low performer" or a superstar.
You cannot be an effective manager and simultaneously maintain enough engineering context to make good technical decisions. If you want to be involved in detailed technical design, you need to switch your career track back to IC. And if you are running a company and you want either good technical decisions or high morale among your engineers, you can't tolerate managers playing at being engineers instead of doing their actual jobs. (With something like a manager maintaining a small low-priority internal tool, or doing a rotation onto an engineering team for a week to refresh their intuitive handle on the state of the codebase & tooling but no expectation of being a net-positive contributor, being the exceptions that prove the rule.)
That being said, in what world does the Eng director look good? Insisting on using SMALLINT (and thus necessitating a change away from SQLite in development) is just terrible, and an Eng director who is halfway competent should realize that such a requirement is patently ridiculous, or at the very least know not to stick their neck out if they know so little about the subject at hand. I'd be furious if my Director of Engineering did stuff like that.
In the end he gets bored with his actual work and lets himself be hijacked by some line function to develop internal tools for them, outside of the purview of any manager. Unsurprisingly he ends up an organizational orphan. Which means no one goes to bat for him to squeeze in several weeks of remote work, which sounds like a pretty outlandish demand to my ears. To think that the fact that some random person once 'approved' it makes any difference just goes to show the naïvety. Obviously he didn't get it in writing from someone able to make that decision, and that's the only thing that matters.