analyze-snap should perhaps compare to non-snap packages

[popey]

While analyze-snap is super useful for comparing the result of changing the compression algorithm from xz to lzo, there is another comparison which could be useful. Compare an xz/lzo snap with the non-snap package startup time. This shows the difference between the snap and non-snap package could highlight further debugging if necessary.

Perhaps optionally as:

etrace analyze-snap vlc --non-snap /usr/bin/vlc

I'd expect the same 20 runs of the application to be done after the xz/lzo runs.