One aspect of finishing which seems to have been omitted in the discussion is that in some cases the finish line is not a fixed, static line at all. We have two main groups of races: in one we start and finish from the club line: an imaginary line between the clubhouse and a flagpole at the other side of the river. All being well, that is fixed and immutable. However we also have series of races in which the finish line is between the committee boat and a buoy. In both cases they are anchored in tidal waters and are susceptible to wind and currents. Both swing about. The finish line swings about between them. Clearly this doesn't have an adverse effect on one design racing since we merely record sequence, but it may on handicap classes/fleets.
On the broader issue of timing, I have it mind to look back through (say) last year's results to see whether our recorders have an inherent bias towards particular seconds: i.e. look at the times which end 0, 1, 2...9 and see if there is a uniform spread, which notionally should be the case. Not sure quite what I do if there is a bias though. Maybe best to keep it as a thought experiment.
Malcolm Clark (scorer at Flushing Sailing Club, and for Falmouth Week)