Yes. These errors are expected and normal.
A fault where the code is trying to connect to iCloud when you have it disabled is perfectly expected, normal, routine. Even things that sound scary or ominous are actually internal code points and have no bearing with the functionality.
I would say, absolutely look at the console before you have a problem so you won't worry so much when you have a specific problem with a specific application or function and then bring that observation with that specific log message to the table in a new question and see if the general advice bears out or if it's indeed something you could learn / fix.
Specifically, in a 10.X.0 initial release, you might expect to see far more of these messages as the new code is still being tested and proved in real life and once the system becomes stable, these debug and support errors are changed to be optional or lower priority. What the developer thinks might be a rare "error" might turn out to happen thousands of times in reality and not be so important to log and certainly not classify as "error"
Here are the counts of errors and faults on my 100% perfectly working, no issues MacBook:
$ log stats
size: 589,735,560 bytes
2,484,914,791 bytes (uncompressed)
start: Sun Sep 10 23:27:15 2017
end: Wed Oct 11 11:03:43 2017
statedump: 6,902
events: [ total log trace signpost ]
[ 36,978,153 33,189,182 18,493 623,275 ]
activity: [ create transition action ]
[ 3,139,162 0 83 ]
log messages: [ default info debug error fault ]
[ 33,127,303 437,500 303 245,043 20,801 ]
ttl: [ 1day 3days 7days 14days 30days ]
[ 623,309 31,524,041 885,274 405,568 399,660 ]
Less than 1% errors and far, far less faults. The sandbox issues a lot of messages when it prevents apps from reading and writing outside their claimed space and that's a good thing IMO.
One very interesting beta tool is Woodpile from Howard Oakley - it seeks to analyze the volume and pattern of messages to help figure out when/if a problem started or ended and might be a very useful tool for people interested in watching their logs.
I resolved this on the same system by installing the 10.13.2 supplemental update, and then the .104 version of the NVIDIA Driver Manager. (the installers are very picky about versions, so pick whichever matches your OS version here; nothing before .104 installed right for me https://www.tonymacx86.com/nvidia-drivers/ )
Restart, and then switch to the NVIDIA Web Driver from the manager:
(edit: driver helped for a while but the memory leak came back. sigh. back to not having scaled mode available on external screens...)
Best Answer
The “stadium lighting” you show is usually a backlight issue where the voltage or lamps are not to spec.
It’s highly unlikely that a thermal situation would cause that, but I don’t want to say there’s no chance or you’re wrong unless I could examine the whole Mac.
100C on many internal parts is just fine, so your reported ranges don’t seem at all out of line - especially on 2016 and newer MacBook Pro.