WARNING: this is a fast fix, i.e., we lost the functionality to use the fstatistic metric, thus all metric computations use the phase metric.

fix semicoherent per segment 2f computation, ignore difference between expected and actually found number of SFTs, always use metric in chain search for the 2nd highe sc and fc stage

In multistage searches use metric / fisher extends for the 2nd and higher semicoherent stages, if metric restricton is not used in stage 1.

Fix SSBprec option, previously disregarded, thus always using Newtonian approximation to compute SSB times.