What has any of this to do with future sensor technology?
Nothing, but it's fun watching them argue about some minute points although they were on the same page.
What they missed in proclaiming that autofocus is not good for anything is that it's actually faster and good enough for most things, and the market is proof of that for the last couple of decades.
Arguing that a camera it's dumb because it doesn't know where to focus it is beyond funny; this not one Asimov's periods yet, of course it doesn't unless you tell it. That's why there are focus points which, shockingly, you can position, or the old, wait for it, focus and recompose.
Manual focus is preferred for macro photography, but that's a pretty small niche, and for some stationary subjects. Or as a backup for the increasingly rare occasion when AF cannot lock and few other very out of ordinary situations.
In terms of accuracy, the phase detect can have 20-30% missed focus which sounds bad (and it is) but you will be able to take more pictures and be faster so in the end it might be a wash, particularly as the MF focus skills will likely be much more variable, some people will be very good and show it can be done as in this thread but many will be quite bad at it.
Now with the increased use of contrast based AF it will be at least as accurate as the best MF. I'm sure the Internet will give enough references.
In the end there will always be situations when AF is not working and that's when you have the backup of MF, but those situations are not the norm, and this days that's what MF focus is for most photographers, a backup and not the first choice.
Paradoxically maybe these days I use more MF for stationary objects because the electronic aids make it easier (eg. Focus peaking, etc)
