Saving tetrode stats should be its own function

[nickponvert]

We currently have a function spikesorting.cluster_many_sessions() which performs two tasks:

• It uses spikesorting.MultipleSessionsToCluster to perform multisession clustering
• It calculates some stats about the clusters found, and saves TetrodeStats.npz files

We should consider separating these two processes, so that the function for calculating tetrode stats could be used on the output of spikesorting.TetrodeToCluster as well. Calculation of tetrode stats, and saving the .npz files, could possibly even become a method of spikesorting.TetrodeToCluster which would then be inherited by spikesorting.MultipleSessionsToCluster.