The fact is no laptop should be getting hotter than 90C. Can the CPU handle it? Yes. Should the CPU have to handle it? No.
100 degrees is problematic and will increase wear on the system. Heat is not good for electronics period. 100 degrees is not normal and not good.
And people seem to be operating under the mistaken belief that if its not shutting off due to heat then its fine and working as intended. The fact is that pretty much no modern laptop shuts off from heat anymore. It just downclocks. If the CPU is running below base frequency under sustained heavy loads then there is a problem; the CPU is not hitting its intended frequencies and thus is not working as intended. Running at base frequency is okay but not ideal. A perfectly working CPU will run at max boost.
I had a dell that would down clock to 1.2 ghz when playing Skyrim (from base 2.0 ghz and boost 2.6 ghz). Did the dell shut off? No. Was it working properly? No. Was there a problem with it? Yes.
Is no one concerned that Apple is selling you an expensive CPU upgrade and then potentially cripples that CPU by not supplying the thermal headroom? Or the appropriate power envelope (85 W for the system isn't enough)?
Since you didn't read the previous discussion either, I'll recap:
• Different programs report different values for the temperature(s), and it's not entirely clear which are "correct" and which are "incorrect."
• iStat Pro and others reported values at 100% CPU utilization around 46 degrees Celsius. However, my suspicion is that some of those lower values are around the heat sink.
• The utility used by the OP shows values closer to 100 degrees Celsius. I poked around, and that's the on-core reported value.
• Here's the kicker with everything you said: having an on-core temperature of 100 degrees is
NOT the same as having a "laptop" temperature of 100 degrees. If the internal temperature is in the 40s or 50s (which it is, even with the CPU and GPU together), then we don't have a problem. You're talking about 100 degrees as if it's some system-wide temperature. It isn't.
Not even close. And again, I can only get 100 degree on-core readings while pushing the CPU to 100% utilization across all cores and with HyperThreading (i.e., 8 processes).
• We don't know if, at maximum load, the CPU is running below base frequency. It's a good bet that Turbo Boost isn't kicking in. Again, this is something that someone running Boot Camp would have to test.
• The last point is a theoretical/hypothetical one. Let's say the laptop were magically doing a better job of cooling (more on that in a minute). In that hypothetical world, would the CPU still be hitting 100 degrees? (It's distinctly possible.) Would it be Turbo Boosting, and if so, how much? There's no way for us to answer these questions. Ostensibly, if there's a fantastic cooling PC laptop out there, we could run an experiment comparing the same 100% utilization on the Mac and the PC and see...but again, that's something someone else would have to do.
TL/DR: I see no problem whatsoever. Could cooling be improved with some Arctic Silver and non glopping it on as if more is better? Absolutely. Does that poor conductivity probably shrink the life of some Apple laptops? I'm sure. Is it really a
"problem" per se? There's no compelling evidence whatsoever to suggest that it is. There's just a bunch of speculation and supposition.