Report 90% credible intervals on distance
Reported by @salvatore.vitale after S190521g
I'm OK doing it for Bayestar, since that assumes the posterior is gaussian (AFIU), but LALInference doesn't make that assumption, and indeed https://ldas-jobs.ligo.caltech.edu/~carl-johan.haster/O3/S190521g/SEOBNRv4_highMassratio_4s_11Hz/1242442967.4472656-333631/V1H1L1_PRELoutput/1Dpdf/dist.png is not a gaussian. Why not give the 90% credible interval?