Here’s an analogy: I have a kitchen drawer that has had its face torn off, because the threads for the screws were destroyed. I tried fixing it twice by drilling them out, gluing a hardwood dowel in, and then redrilling for screw threads.

After the second failure, I asked my FIL - who is a woodworker - if I should bore them out to a larger size, so I could use a larger dowel with more surface area for glue, etc. He said, “you could, but the failure point is the threads, because you’re driving parallel into end grain. A hardwood plug inserted such that you’re driving into edge grain would hold. Or, you could just move the screws, and optionally plug the old holes for aesthetics.”

I was vastly over-complicating something that had an extremely easy solution. This problem exists everywhere in tech; people will recreate existing technology (usually in a worse fashion), or create Byzantine pipelines for a problem trivially solved by a bash script, etc.

If you consider the available options for something, and then decide, that’s one thing. If you make what is objectively the wrong choice, that still might be understandable - maybe you want to learn something new, maybe it doesn’t matter at your scale (which tbf is true here, though Vercel’s pricing might cause pain if the site exploded in popularity), etc. But the point is, you should understand trade-offs, and what already exists.