OK, so if that is the case, what is happening here is that a 'sample' of the overall oil is rising in temperature during operation of the retarder and then when the retarder is 'disengaged,' the sample is released (dumb way of saying it, but you get it) back into the overall oil flow, so the apparent oil temperature decreases?

If that is the case, then it's no big deal? Oil certainly can be be as hot as 220F without destroying its lubricity.