The primary objective of a slower print speed for the first one to two layers is to ensure a bonding to the build plate. Print quality is not likely to improve at speeds slower than the stick/no-stick speed. For a given brand and color of filament, there will be a binary assessment. "This speed sticks to the bed" versus "this speed doesn't stick to the bed."
If there are quality variations in the print at different speeds, there may be other factors involved, such as extrusion rate, retraction distance, z-hop, etc.
It's difficult to consider a diminishing returns reference, as it's going to be "it works" or "it doesn't work." The increase in print time for going slower isn't going to contain much value in this situation.
To determine a specific speed requires experimentation. The default for your particular slicer in use is the starting point. One can increase the first layer speed by 5 mm / second and run a test. Examine the results for quality and adhesion. Increase by five for each succeeding test until the results become unacceptable.
Note that different plastics will have different speeds. I've not experimented with changing the default speeds except for TPU, which requires slower speeds.
You may discover that you can get away with much higher speeds than the default with one particular filament brand and color, but changing brand or color may change the results.
To address your comment, the two sections of your question appear to be opposites, but for both, only experimentation will find the specific range (not exact speed) you seek.