There's no one right answer to this. To some extent, larger pixels are better due to their light sensitivity. But, there are also benefits to shooting at a higher resolution and then downsampling to hide some of the imperfections. Where the balance point is depends a lot on the specifics of the situation and the priorities of the camera manufacturer

Thank you so much for your reply! I had a follow up question too - has been bothering me for a while - given what we know about sensors and pixel pitch, how does a newer sensor with a specific pixel pitch improve upon an older sensor with the exact same specs. Does it then all boil down to how the on-board processor process those pixels?

It is an interesting question, and one I have thought about myself. ...Four noisier pixels taking up the same space as one pixel that is less noisy... When viewed at the same distance, how do they compare? The ironic part is that most cell phones are only 1080p anyway, which is around 2 megapixels. A 4k cell phone screen would be 8 megapixels. So assuming no extreme zooming, multiple pixels in an image will be binned together into one pixel on the screen anyway. To me personally, the bigger downside to so many megapixels is simply that they (often needlessly) take up a lot of storage on your device for very little benefit (even assuming no issues with noise). I do professional photography part-time, and clients give very positive reviews, and I have a 24-megapixel full-frame camera (a7 III). The person who made this very video filmed this filmed it at 4k (8 megapixels) on an a6400, which also has 24 megapixels (and a somewhat smaller sensor than my camera - APS-C)... As resolutions increase, the resolution becomes less important, and other things like lens quality account for more of the end photo quality. (In other words, a 12-megapixel camera with a super-sharp lens can yield a better picture than a 100-megapixel camera using a lens that lacks sharpness and clarity. Even though I feel that having a lot of megapixels can have downsides, I would also like to see how the noise can be controlled when four noisy pixels take up the same size on a screen as one less noisy pixel.

Yeah, it's a concept called Pixel binning. Back in the CCD days this could be done in hardware, but now it's a software only thing. The downsampling method that you use to get to that level is also very important. If you have a high resolution sensor, and it uses, let's say line skipping to achieve a certain resolution for video, you've essentially cut out the light you aren't using, bringing more problematic noise performance than you might expect from pixel binning, or oversampling.