Choice of strain interpolation order in Stochastic sources

Hi @j2b.bayle , I think it would be useful to be able to choose the strain interpolation order for stochastic point sources and stochastic backgrounds, as it can drive computational complexity and memory demand. Are you okay if I go ahead with that in a commit? In the future, we may also want to think about a proper way to "interpolate" noise (polynomial interpolation is not statistically correct). I have a few ideas for that.