Osama and George W. – an example of lineage-branching
There is a difference between genealogical branch points and mutation points, when it comes to examining a YDNA-based tree.
A genealogical branch point is a place where two sons are fathered, and henceforth they each have established their own paternal lines. It does not mean that a mutation has necessarily occurred at the branch point.
We offer the following, tongue-in-cheek example. Abraham sired Ishmael and Isaac. That's a branch point, and there are subsequently two paternal lines. There'd be plenty more branch-points after that, I'm sure, but for the sake of this example, I want to ignore them. Let's suppose no mutations occurred over the eons until George Herbert Walker fathered George W. "Bingo!" Mutation! If FTDNA were to test George and Osama (good luck locating him!), they'd find that one-marker difference -- but there'd be no way to tell where, between Abraham and George W., it occurred.
All one can say with certainty is that a branch point occurred at or before the mutation point that differentiates two individuals. How far back? Who knows! You can even have relatively recent branch points which, if no mutation has yet occurred thereafter, have exactly the same marker-patterns.
In the "Gossett Model" We've presented, we show the paternal lines that can be differentiated by markers, and the positions (relative to each others' branch points) of these differentiable paternal lines. Realize that we can't say with certainty how many generations ago they occurred. The spread in confidence intervals is simply too large, especially when relatively few generations are involved. Consider a coin-flipping analogy: Over the long run, we expect 50% heads and 50% tails, but when one deals with only a few flips of the coin, it's unreasonable to expect equal numbers of heads and tails. Among our project participants, we see some whose lines have experienced one marker-change since the "patriarch," while over the same span of generations, other lines have experienced as many as four marker-changes.
That is why we see little justification for presenting a time-scale with the YDNA-based trees – unless coupled with a paper-trail in which we have confidence to assist us in more precisely fixing the branch points.