I'll have to look back in the threads for where this was discussed earlier, but Martin reminded me that one way he looks at this is with a "comfortable temperature" analogy, with pleasure in the role of temperature.
Once a comfortable temperature is reached, the temperature does not get more or less comfortable by changing, or remaining in place for a longer period of time.
In this analysis, that would mean that the question of "how long to remain" at a comfortable temperature (whether it is better to remain at a comfortable temperature for a longer time) would need to be answered by other considerations.