So, your two arguments are "Some people don't follow it, so no one should." and "He doesn't follow it already, so no point in correcting him."
Cool.
"No" on both counts, but I'll just take a different approach.
Semantic versioning does not translate perfectly into game development. Let's start with the major number, 1.0.0, "breaking changes." Does this mean every time the game makes a change that breaks old game saves, we should increment the first number?
Here on f95zone, there's a widely held sentiment that 1.0 should represent a finished game, while 0.1 should be the first draft. If we followed Semantic Versioning rigidly, then adding a tiny new variable to the game that breaks old saves in 0.1 would immediately jump us up to 1.0. This is a bad idea.
To your credit, you said "overhauls OR massive expansion of code." However, if we're on version 0.9, and the next minor update completes the game, you can bet people are gonna wanna see it labeled as 1.0, with further updates becoming 1.1+, not 0.9.9.9 into perpetuity just because the dev never releases massive updates.
In other words, it's fine to use the first number as a marker for milestones achieved. Which is how it is done, both by indie devs and AAA game studios alike!
To be fair, the third number is typically reserved for bug fixes and general polish, so that holds most of its precision from the original concept of Semantic Versioning, even in game dev.
The middle number is super vaguely used. Really just any normal update that adds more content. This more-or-less holds with SV anyway.
Programmers love to argue what the "correct" way of doing things is (see: tab-spacing debates). I am telling you, do some research on SV specifically in the context of game dev. I think you'll be surprised how widely the vague notion of "big.medium.small" is followed across the board. And yeah, this can be interpreted as them following SV, but it's a looser definition than you make it out to be.
tl;dr: Relax. It's not a big deal. "Big.Medium.Small" is not that far off to begin with.