And diffraction limits kick in much earlier with such high pixel densities. OK if all you want to do is shoot wide open - but for the rest of us shooting wildlife (so not just small BIF) with long teles we often need to close down to give our subject the Depth of Focus we need to bring more than just a small part of the head in focus .. and light/motion speed permitting this is what I do.
"Diffraction kicks in" perhaps ideally needs putting in some context.
The laws of optics (which I find to be accurate) are that if MP is increased - there is more image resolution at every aperture -
including diffraction limited apertures.
What is going on is that
image resolution never reaches 100% of lens resolution (measured without a body)) or sensor resolution (measured without a lens) - and there is always some surplus resolution that can be got out of a lens even at
image resolution diffraction limited apertures - because the mathematical diffraction limit is unlikely to be reached.
Although MTF 50 is probably not a good indicator of the human perception of sharpness, it is a good indicator of when diffraction starts to reduce
potential image resolution at 50% contrast with a specific MP sensor.
Many lens aberrations that limit resolution at wide aperture reduce on stopping down - partly offsetting the negative effect of diffraction at smaller apertures.
This is why despite there being a diffraction limit at f1.2 other aberrations prevent the f1.2 diffraction being reached - and image resolution increases by f2 - despite the negative effect of diffraction being being greater than at f1.2.
45 MP has close to 33% more pixels on the long dimension of the frame than 24 MP - with the potential of a 33% greater
contribution to image resolution from the sensor - though combined with the lens resolution separate contribution to image resolution - image resolution does not go up by the full 33% even where diffraction is not a factor.
Taking 45 MP combined with a high resolving lens as a starting point - it is probable the increase in resolution would start to be squeezed at smaller than f5.6 if there was a DX 45 MP body, the increase can start to be squeezed (depending on the lens) at smaller than f8 on FX, and on 45 MP digital medium format the increase should not start to be squeezed until smaller than f11.
In theory if f16 depth of field is needed on FX format medium format images should be less diffraction limited at f16 - except about f22 with a greater diffraction squeezed aperture is needed to get FX f16 depth of field on digital medium fosqeeze
A practicality is that the amount of potential extra image resolution increase from higher MP sensors gets squeezed - but is still in the plus zone despite the diffraction squeezing effect - definitely down to f22 and potentially f32.
Few would refuse to use an f1.2 lens at f1.2 if they wanted f1.2 limited depth of field - despite image resolution being higher by f2.
A current developing internet myth seems to be shooting at f16 with 45 MP is "the end of the world" - despite image resolution still being distinctly higher than shooting at f16 on a 24 MP camera.
A separate topic is that higher pixel density at higher ISO's does generally reduce image resolution, increases noise and reduces dynamic range. Even so many at least some of the time used 3,200 ISO - before denoise software became readily available
