If you meant Onondaga's warning, it was for pressure rise from touching the lands. That's usually on the order of 20%. It doesn't just jump up to that number when contact occurs. Dr. Lloyd Brownell showed back in the 1960's that it actually starts increasing before then as the escape path for bypass gas is cut off by it narrowing the annular opening between the bullet and throat. I figure around -0.020" to -0.030" from the throat usually eliminates it with most pointed bullets. Brownell showed the pressure minimus for a round nose bullet didn't occur until the bullet was a whole quarter inch back from the lands (see plot below).
The amount of the lands for best accuracy depends on the chamber and bullet. For longer nose VLD bullets, Berger found some guns needed as much as 0.150" jump for best accuracy. For shorter ogives, it is generally less. Numbers like Onondaga gave if 0,005" to 0.020" are commonly mentioned, but you should explore well outside that. Writing in the 1995 Precision Shooting Reloading Guide, the late Dan Hackett, a competitive benchrest shooter, described a 220 Swift rifle that would not group tighter than about 0.5" at 100 yards on average (5-shot groups that ranged from 0.375" to 0.700"). He always loaded bullets to jump 0.020" to the throat. Then one day he accidentally turned his micrometer seating die the wrong way for a bullet change and wound up with 20 rounds loaded to jump 0.050" before he noticed the error. He considered pulling them and reseating them, but decided to save time and just shoot them for practice, not expecting much accuracy. To his astonishment, those 20 rounds gave him two 0.25" groups and two true bugholes in the 1's (what benchresters call groups over 0.100" and under 0.200"). So, you don't really know what will be most accurate until you try.
Rotating the cartridge as you load it to keep the bullet straight is a technique that has been around for a long time. However, I never got it to work very well. It helped a little, but not enough. Two more effective approaches:
One is to use a Lyman M die to put a seating step in the case mouth to set the bullet in so it is straight up and down when it first contacts the seater. When they start in straight they tend to stay straight. Depending on the dimensions of your chamber, you may or may not need to iron the step out with the crimp shoulder of the seating die, but if you don't need to, just leave it to act as a spacer to help center the neck in the chamber.
The other is to use a high-end seating die that will tend to minimize and even correct straight seating error. The Redding Competition Seating Die is the best one I've found so far. My results with it echo
German Salazar's findings.