Status of variable sample-rate fitting?

[moustachio-belvedere]

Creating this issue to track status of variable sample-rate fitting changes. My (vague) memory of our last group discussions on this is that we should consider getting rid of variable sample rate fitting due its inaccuracy and slowness, replacing it with coercion of user data to constant sample rate.

First decision to make is whether we get rid of it completely or we just have more warnings like:
https://github.com/JuliaRheology/RHEOS.jl/blob/master/src/base.jl#L574
in other relevant places and the documentation.