The shoe homologation aspect is an interesting one.
At one end of the spectrum you have someone running barefoot. No cushioning, no extra "bounce" or "spring".
At the other end you have the theoretical case of someone using large springs attached to the bottom of the shoes (or integrated into the shoes), and probably making the average person able to run faster than 13mph for 2 hours.
All running shoes will be somewhere between these two extremes. Some offer more bounce/response/spring than others. The IAAF will have drawn the line somewhere.
What the IAAF allows in order for a shoe to be homologated will not be found in a vague two line snippet in its rules about shoes. There's probably going to be some big technical document in a dusty filing cabinet in the IAAF head offices that maybe Nike are aware of and have a copy of and knew how to push the limits of.
Those shoes will probably be tested at some point, or are already undergoing testing, by the IAAF. They'll either be found to be ok, and loads of people will want to wear them as they will be homologated, or the IAAF will kick up a fuss and say they offer too much "bounce/spring" and no-one will be allowed to use them in a standard race.
Either way, Kipchoge ran a marathon in under 2 hours on a near flat closed (to people and other competitors) course, with rotating pacemakers, laser pacing, in shoes that may or may not get the IAAF stamp of approval, and probably even some mild benefit from a car 15 yards or so in front of him (physics says it would be negligible given the pacemaker formation in front of him).
There are a lot of caveats there but, now that this has happened, I don't think it will take long before it's done again under normal marathon conditions (open event, no rotating pacemakers, no car to potential aid the draft, etc).
There are also other ways to run faster "marathons" that don't comply with the same rules:-
* Running down a big mountain would give you a really good advantage, but the marathon record rules prohibit excessive elevation loss between start and finish.
* Canicross could also give a good marathon runner a really good chance (I believe there are some sub 13-minute canicross parkrun times out there and one runner+dog combo was closing in on a sub-12 minute parkrun[1]
* etc
but, of course, these do not qualify as the "marathon record" since that has a bunch of stipulations which were ignored for the official record (some of which are entirely arbitrary too, such as the maximum amount of elevation loss).
1.
https://www.fastrunning.com/features/ben-robinson-aims-sub-12-minute-5k-help-blake/9867