update spectrogram code to be more consistent

we now have a couple tf-decomposition functions we wrote (`calc_mt_tfr` and `calc_cwt_tfr`) as well as one from bmi3d that works in a completely different way (`multitaper_lfp_bandpower`).

a wrapper around `calc_mt_tfr` and `calc_cwt_tfr` to average across frequencies would be more consistent with the way `multitaper_lfp_bandpower` works. we should update the naming to be consistent, too.