The camera's processing is what it is, unless you output raw and process it yourself (which some camera apps allow, some don't, but it's a much slower workflow so wouldn't be my first choice with a phone). But I'm not even thinking about quality of interpolation: handling of noise, tone, saturation, contrast, not over-sharpening, these things matter more for the image than simple pixel count. All phones use small sensors and are increasingly reliant on computational photography techniques, such as taking many underexposed pictures rapidly and combining them (stuff which you, as the user, never see going on). And spec sheets don't show any of this stuff. This is why I say that the way to judge a phone camera is to look at the pictures it takes and decide whether they are good enough (independent reviews or photos taken by users, not samples produced by the company themselves - more than one company has passed-off photos taken using professional cameras as phone samples).
But as a general point, you can never get 64MP images from a 48MP camera or whatever. Sure, there is upscaling software of increasing sophistication out there (the more sophisticated tools cost around $100 for a license, and require a real computer rather than a phone), but you can't create information this way, it's still just using a mixture of interpolation and using learned patterns to insert guesses at the detail that isn't recorded.
More seriously, don't assume that higher resolution = better camera. Most of these "64 MP" cameras produce 16MP output by default (pixel-binning, 4 pixels -> 1), while "48 MP" defaults to 12 MP. And while they generally have a mode that can output the full resolution, don't use it in anything other than bright light and don't assume that it will resolve as much extra detail as it sounds (I've looked at samples, such as in reviews on gsmarena.com, and often low-contrast detail, such as foliage, is smeared away by the noise you get with many small pixels, each of which is not collecting much light). And if the lens is not sharp it won't matter how high resolution the sensor is. So as I say, I judge phone cameras by the images they produce and not from specs.
(Also, as a final comment on phone specs, remember that a 64MP sensor only delivers 15% more resolution than a 48MP sensor, which is about the limit of what is perceivable when looking at an image of a reasonably large screen, and probably completely invisible when looking on a phone screen. This is because resolution improves as the square root of the number of pixels. And at this level differences in lens quality will matter far more than differences in pixel count).
I think I can't really advise on the last points because your usage is so different from mine. I hardly record video at all, so sequential write speed doesn't matter to me, and hence I have no reason to know whether these different things make a difference in reality - though UHS 2.1 is old technology, so presumably that's the manufacturer cutting costs. As for battery, I'd try to find reviews which test the actual battery life of the device rather than just look at the capacity: your table shows a longer battery life for the Poco despite a smaller battery, which can happen if the components are more efficient, but the question is how did they test the battery life (i.e. does their test reflect what you would use it for)? So maybe find some other reviews of these devices and see whether they agree about which device has the best battery life (e.g. GSMArena's reviews report a better battery life for the M51 than the X3 Pro, which makes sense for a larger battery and a more efficient processor). But also try to judge what you want in terms of battery life, and how that compares to other factors: unless you are like me, where one factor trumps all others to a huge extent, it will be a balance of factors, in which case "what is good enough?" is as important as "what is best?".