To me, it makes sense to count from the flip, because you then document a major change to the plant's environment that you created. Then you create data. If you end up running the same clones from the same mother for a few runs, under identical conditions it seems like they should finish really close to the same time, unless you alter your criteria for "doneness". The number one rule of all gardening is to pay attention to the plant! Accumulating data can be a great way to pay attention.