I don't think the "folk process" can easily be measured in terms of the number of changes, the number of singers, or a period of time. Like biological evolution, it is easier to observe with hindsight rather than in real time.

Every singer makes both conscious and unconscious changes to a song. Some of these may be substantial - adding or leaving out verses, changing the order of verses, improving the text or the tune to make them better to sing. Others may simply occur during the process of performing the song over a period of time.

One individual's changes aren't in themselves sufficient to turn the piece into a folk song. However, if enough singers take up those variations, add their own, and pass them on, in time we will find a separate version of the song which is clearly distinguishable from the original. Along the way, it may be difficult to distinguish the long-term variations from individual interpretation.

An example is McColl's "Dirty Old Town". Nowadays, it seems to be sung to a 4/4 tune, which has the effect of dragging out some of the words interminably: "I met my looooooooove ....". In the recording I have of the man himself, he is singing in what appears to be a 6/4 metre, which fits the words much better and shortens these long phrases. However the 4/4 version now seems to be firmly established, to the extent that it has probably taken over from the original. That is the folk process at work.