Improvement: make option for energy binning

[aidancrilly]

Currently the NeSST spectrum is defined at energy values, i.e. the spectrum is defined as a series of spikes. This can lead to some undesirable behaviour - spikes in scattered spectrum especially for restricted rhoL angular ranges.

An improvement would be to make the spectrum calculated within energy bins (assumed constant within bins).

This is linked to #14

Making this an option is perhaps challenging but would allow backwards compatibility