24 mp might be a bit too conservative even for the 1Dx line, 28-30 mp might be more likely. Autofocus should be where they concentrate their improvements (I expect to see an improved autofocus introduced with the 7dIII, as well) The market for the 1Dx series is conservative -- pros who want rock solid reliability and no surprises. Improvements will be incremental. They will add full touch screen, wifi and up the frame rate a notch. This won't be the last 1Dx DSLR, but there may be a 1Dx R introduced in about three years that will be sold alongside the DSLR for at least another generation. Pro 1Dx users will likely be the last group to move to mirrorless (if indeed they ever do). Canon will keep making and selling DSLRs as long as the demand is there.
I think the number of people who would feel a +2 fps upgrade to the new 1D model was worthwhile is *much* smaller than the number of people who would put up money for one if their was a +5-10 megapixel upgrade. The frame rate is pretty good - better than needed for most uses. The resolution is not so good for a number of applications, to the point where the 1D isn't considered the best Canon camera for some people. But I see no evidence that Canon has the capability of putting through more than 280 megapixels per second (14/sec@20mp) while using three Digics in one ginormous body, and I doubt they would lower the maximum fps at full resolution. My guess is they have the capability of giving about 10 fps at 30 mp and would require the 1D form factor to give it enough of a heat and power envelope. Just don't see them doing it. It also appears that to do the equivalent in mirrorless leaves only a portion of the required processing capacity available. A 1R might do 8fps at 30mp (in servo, which is about 3 times as fast as the current R in servo) and need to be as large as the DSLR variant.
If they came out with either such increased megapixel upgrade, I'd definitely buy. But if they came out with a DSLR with +2 fps and some of the obvious upgrades, like the touch screen the 2dx2 inexplicably lacked at launch, etc., I'd buy a used 1dx2 after they came down to $2.5-2.8k used if I wanted to stay with Canon, or - like as not - look seriously at alternatives.
I'm in limbo now. I bought an R yesterday, and it's interesting, but annoying in some ways. I was surprised how much less usable the touch screen focus point manipulation is on the R versus the M5. I was expecting it to be at least as fast as the M5 I once owned. I bought the R to stick a 50 1.2 R on it, and that's impressive, but I'm not sure I'll keep them if it doesn't get easier as I learn to use it better. Images are just like those from my 5D4s, not surprisingly, so that's adequate. It is, however, unacceptably slow in the frame rate department.
All this makes me think that if Canon doesn't prove it's ability to throw lots of megapixels (at least 30) through to a memory card more than 8-10 times per second in a mirrorless body, then it doesn't make sense for me to invest in EF glass (on the way out) nor R bodies (not capable enough). If the 2020 Olympics offering is a 1 series body upgrade about as incremental as the last one, I won't expect Canon will provide what I need, which would greatly sadden me. If they came out with a great 1 series upgrade, I might get that as a stop-gap to allow me to use current glass and give time for mirrorless to catch up. Not that other people's use cases aren't well taken care of by these current offerings. And not that the RF lenses are the most exciting thing happening in the industry right now from an academic perspective. Between now and that next release, I'm experimenting with both the R system, and I'm going to also play around with either the L mount stuff - once Sigma's glass comes out - or the next rev of the Sony A9 or A7r. Best case: Canon surprises us with something great. I have my fingers crossed, but I'm worried enough that I'm spending time and money exploring other things to do due diligence.