Intel is exploiting their 14nm++++ process to an extent where they just pump power into these chips making them hot. their next node shrink will - hopefully - see a temperature decrease.
It's not intel pumping in the power, intel has pretty tame guidelines, which although still higher than ZEN are not all that crazy,if you follow intel's guidelines the 10900k should draw pl2 of 250W maximum for a maximum of 56sec which makes sense for workloads that the user has to sit in front of the computer and wait for,if you have to wait for 2-3 minutes for something to finish you will be glad if the CPU can use twice it's power budget to finish it in half the time and since it goes back to TDP afterwards the cooling can slowly dissipate that excess over time even if you only have a TDP cooler.
If you do something that takes 10hours (or a week) it will stay at PL1 of 125W for the whole duration except for the first minute,benchmarks will still show a max draw of 250W but over the whole duration of your workload it's going to be 125W exactly.
It's the mobo makers that take those defaults and increase them to infinity to look better against some other mobo,anadtech did a pretty good round up on it.
https://www.anandtech.com/show/13544/why-intel-processors-draw-more-power-than-expected-tdp-turbo
Also the same thing happened to ZEN although with a different technique.
ZEN depends on info from the mobo to determine how much power they draw and weather they should keep boosting or not and mobo makers take advantage of this by just saying that the power draw is lower than it actually is,the stilt found it out and you can see by examples how the readings from HWinfo actually show lower package power draw by messing with the setting.
https://www.hwinfo.com/forum/thread...er-reporting-deviation-metric-in-hwinfo.6456/
Bottom line temp problems are almost always down to mobos screwing with settings to look better.
For intel they have to increase TDP to make them faster for ZEN they have to decrease TDP to make them faster.