20mp and 25mp MFT are effectively as good as "40mp" xtrans APS-C at everything but computer generated noise charts, because of xtrans. On the most academic level only 25mp should be close but not exact, but if not pixel peeping at 10x zoom the most academic level doesn't exist. Compared to lower end xtrans cameras that don't have the 40mp meme, 20mp MFT blows past it.
Monochrome and foveon sensors are 50% higher resolution than bayer sensors.
2x2 RGBG and RGBW bayer sensors are 50% higher resolution than xtrans sensors and 100% higher resolution than quad bayer sensors (quad bayer cameras like sony a7siii actually have more megapixels than they say, but they're grouped in 2x2 blocks of each color instead of 2x2 blocks of all 3 colors`)
When you have more megapixels and a larger image, you don't necessary have more details. Megapixels are image size, what resolution they communicate is a complex technical subject that would require essays about color pairs and line directions to explain. You can see how fuji loses detail resolution and has color resolution issues here:
https://petapixel.com/2017/01/27/x-trans-promise-problem/It's just a fact of digital gain structure. Except for foveon and monochrome, which work more like film, megapixels aren't a direct statement of resolution. The actual resolution varies wildly depending on the color of what you are photographing and how it lines up with the blocks because each pixel is only one color and the computer is guessing at the other 2.
All of this is assuming you will actually use your megapixels. The print size equivalent of viewing crops on dpreview is massive - several feet wide. If you are not going to print several feet wide, every camera will look identical once you process your images properly and you're just wasting money spending more for supposed image quality
