I know this thread is not about 4K video processing and we are off topic, so my apology.Oversampling is the way to go. You get even more detail than 1:1 4K. It's the 4K method Z6/A7III/S1employ.
"Oversampling is the way to go" is not necessarily the way to go, I think. In signal processing theory, oversampling (i.e. getting more samples from a signal, as in Nyquest theorem) introduces unwanted artifacts and aliasing and these two need to be filtered with a Low Pass filter, which is also known as AA filter. So you may end up with better noise control, at the expense of slightly more blurred picture. Doing so needs more processing power manifested by reduced battery life and more heat generated.
So an engineering decision for camera manufacturers is reaching a compromise between the sample size, crop factor, sensor read-out speed, processing power and heat management. Some manufacturers prefer to go for oversampling as you mentioned, the others may go for pixel binning (Z7?), line skipping (A7Riii) and cropping (EOS-R) or a combination of them. So far, no camera manufacturer has come up with a technology which is a clear winner here.