Is there any way to rarefying the reads to certain depth instead of to the lowest library count ?

This is certainly possible to add such an option. Can you tell more about the use case?

Some samples may contain a significantly low reads and some may have a very high reads. Say a range between 10 and 10000 reads. Rarefying at certain depth would be more reasonable although some sample with a very low reads will have to be discarded eventually ( below the depth cut-off).