VIC-20 keyboard seems to have become sort of a standard for [...] the MAX Machine (in membrane form), C64, C16 and (with additional keys added) the 128.
The C64 decision was easy, as it was done to save development time - existing part. Similar for the C16. For Max as well for the C128 it was a compatibility issue.
What keyboards and layouts did ...
Another precision test, previously mentioned in a comment by Tim Locke on this site, was published in Antic magazine Vol 1 No.4. It was submitted by one "R. Broucke" (possibly the late Roger A. Broucke, then at UT Austin):
30 FOR N=1 TO 1000
60 NEXT N
70 PRINT S,X
90 PRINT "CORRECT RESULT: 503.54380215, 1.23"
As Raffzahn points out, unfortunately you can't "adjust" the OS; you
need to replace KERNAL code to have a different number of screen
columns work mostly as expected.
So while there are demos out there such as Big One (s).prg and
Big 2.prg (PAL version), which are nice to look at, they
don't help when you actually want to program using such a screen.
Is it possible to adjust the actual value for rows and columns for the OS without the need for reimplementing the CHROUT function?
No, you need to write a 'driver' and have it hooked.
While the Kernal does have attempts to support size independent operation, like with the SCREEN call at $FFED, returning screen size in X/Y to ...