The entire premise was that (according to lolfail9001), motherboard manufacturers are quoting a 4.3 GHz limit solely because of thermal reasons. Except, this can't be true, because then you could run at 4.8 or whatever the "real" stable frequency is for some minutes until the heat builds up.
You were the one claiming a 4.3 OC was some hard limit:
You can see in the XTU screenshot that they are just increasing the allowed turbo frequency at various core counts. If the CPU were electrically stable at higher frequencies (i.e. no voltage limit), they could have set a higher target frequency and let thermal throttling handle it. Since they didn't, this implies that 4.3 GHz is a hard limit, regardless of cooling.
My entire point of mentioning cooler temps mean high freq with same voltage was to dispute your theory of 'thermal throttling' which is an absurd way of achieving a stable overclock. if it was even possible. Most if not all cpu oc software, or bios-es,do not let you tweak a thermal throttle - its typically a thermal shutdown limit - hit it and your computer shuts down.
Your OC is so amazing that it can happen with the fan off?
Not the OC, the cooling system. You also didnt say fan off, just tweaked to min %. Regardless, I could probably shut off both my rad fans and do a 1min run of p95, or ryzen blender, or a cinebench MT run, due to the specific heat of water, along with the mass of the rest of the cooling system.
Looking at the pic of the 4.3 Ghz OC, thats probably the low end of whats possible. No temp issues, voltage is low for intel 14nm (maxxed voltage would indicate a wall, not low voltages). I'd say those are just quick and simple OC's, no real attempt to find the true limit. A 4.5 all core with a good cooling system doesn't seem unreasonable, esp since it can already do it for 2 cores, so its not a process or voltage limit.