Hatte ich mir jedenfalls mal einst abgespeichert:
In most cases for a given data rate, a 44.1 kHz mp3 will be of higher quality than a 48 kHz mp3.
Here's why:
Using a very simplified example, if you encode a 48 kHz wav file to a 48 kHz mp3 file at 48kbps, each sample in the wav file gets allocated 1 bit.
If you encode a 44.1 kHz wav file into a 44.1 kHz mp3 at 48kbps (same data rate), each sample in the wav file gets allocated about 1.1 bits.
Depending on the mp3 encoder, there may also be different bandwidth settings when encoding at different sample rates. For example, when encoding at 48 kHz, the bandwidth setting could be as high as 24 kHz and when encoding at 44.1 kHz the bandwidth setting could be as high as 22.05 kHz. This means the bits pool used to encode the 48 kHz file has to cover a wider frequency spectrum than that same bit pool used for the 44.1 kHz file. Because the bits in the 48 kHz file need to be spread out over a larger frequency range, there is a decrease the overall sound quality.