This is true, but the difference is splitting hairs. E.g. 80+ Bronze is most efficient at 50% load (85% efficient) but only 4% less at 20% or 100% loads (81% efficient).
I argue that these days since most PSUs will spend most of their time at very low loads (<50W) people should be getting the smallest PSUs they can. E.g. a 1 KW PSU at idle might only be running at 5-10% load.
Either that or get a Platinum PSU which has a 90%+ rating at 10% load.
I appreciate car analogies are traditional, but they're also usually wrong, like this one.
Sustained high revs are bad for typical engines because (AFAIK) the oil circulation often isn't sufficient at the high end. To state the obvious, there's nothing really like that in a PSU.
A better analogy would be a bridge. A bridge might be rated for 30 tonnes, in which case you can happily drive a 30 tonne truck back and forth over it all day long without expecting it to collapse.
This is because the engineers already built in all the safety factors to the calculation, so you know 30 tonnes is safe day in day out.
If your way of thinking were applied to bridges you'd see a 30 tonne weight limit and only ever drive 15 tonnes over it to avoid stressing it.
Electronic components are made in the same way as a bridge, individual components and the whole unit have their own tolerances and limits, and (if designed properly) are kept well below them, even at the rating.
Now that's not to say running a PSU at 100% 24/7 isn't going to shorten its lifetime compared to using it sparingly (it will), but it shouldn't be much shorter than running it at 50% 24/7. But that's a different matter to a PSU going bang.