Never used OBT, I don't think it's helpful. The only way to get statistically relevant data is to shoot a sufficient sample. A small sample may indicate a poor performing load but the converse is not true ie a small good group is not indicative - or more precisely, not sufficient to draw a conclusion from. The sixth shot of a five shot group may well open up that otherwise tight group...
I have read that for each load increment a 5 shot group, repeated 5 times on separate targets, and then overlaid will give the best indication of optimised loads. ( 5 targets so as to avoid rounds going unmarked through a ragged hole).
Or: shoot some five shot groups and take the best ! It's unlikely to be the optimum load but it may well be good enough !