The different perspectives all support this. The simplest would be length contraction.

Say you stand on the surface, unmoving. A locomotive of length 10m is moving at 99% c. To you it appears as a length of 1.41m (correcting for light effects, or assuming the train is travelling parallel to you).
Your reference frame is correct, so the probability of being hit is 7x less than for a still train.

However this only works for a flat train, since a real train sweeps a volume too. If the near speed of light train has height 2m, and meteorites travel 200m/s (accurate for all but huge ones), it can move on a roundtrip of length 3000km and run into any meteorite falling on it before it hits the ground.

The old minutephysics video still comes to the conclusion speed doesn't matter, because it calculates with a set distance. If that is the case, the same volume is swept forwards no matter the speed, the only variable is how long rain falls in from the top so faster is better.
If the time out stays constant though, then faster is much worse.

For the factorio case, my guess would be the meteorite only interacts on impact, so the train is effectively height 0 and speed is indeed irrelevant until lwngth contraction kicks in to make a hit less likely. Of course faster trains might not induce enough demand to keep them moving, so the meteoride may hit a still train which is less cool and also restores the chances of a hit.

For coolness, clearly the train should be moving just fast enough to never sit idle.

Meteors aren't generally aimed, so barring obstacles in its path, they'll hit just about anywhere.

I'm sure there's some higher or lower odds based on hemisphere and latitude and the typical trajectories of meteors before they get to the planet, as well as relative density of the atmosphere... but that's going to affect general areas, not specific ones. Whether or not you're moving and at what speed (unless it's fast enough to move between continents or something 🤪) won't matter as much.

Mind you, that's based entirely on assumption and no actual fact. I'm sure a person with actual accredited smarts would be able to dispute it effectively.

I was talking about real life, but I think factorio calculates damage in more than one physics iteration. If that's the case, your odds increase a very small bit by going faster, but a bit that is way more significant than the "indistinguishable from zero" you get in real life.

Oh, and just to add, your hits of being hit by a meteor on any single trip reduce the faster you move. But the trains are always out there, so their odds of being hit in a game-play don't change (or change very little).