There is no need to soak in ice water. Running cold tap water for 30 seconds will drop the blade to room temperature. If you can hold it in your hand and it isn't warm, it is ready to go back in the oven.
With 52100, using two hour long soaks is a good. Simpler steels do fine with one hour soaks.
What can happen during a slow cooling is that any RA could stabilize. There is some issue with the newly created untempered martensite not being stable, but the RA is mostly it.
This is really more of a theoretical metallurgical discussion that a real world concern. As I regularly state, you will never be able to tell the difference. A laboratory might be able to tell....might! As far as blade quality and hardness, it won't matter if you cool it slow for an hour, wait overnight, or plunge it into ice water, between the tempers.
We are talking about single digit RA% in most carbon knife steels, and the cooling rate between tempers making a difference of a few hundredths of a % increase of that small percentage of RA. Without an x-ray diffraction or SEM in your shop, it would be hard for anyone to accurately detect.
As an example, lets assume that a blade had 10% RA ( which is probably much higher than it really would be if properly quenched). If a slow cooling rate between tempers increases that 10%, the final Ra would now be 11%. In actuality, it is more likely the difference would be 5% vs 5.05%. And even this small amount of RA would be reduced in the second temper.
On steels with high RA amounts, a third temper is useful, but there is a diminishing return after the second temper.