Sorry @scd. You are missing the point entirely.
That paper is doing something different. It is looking at naive estimates (e.g. assuming constant mutation rate) over a much larger timeline (e.g. 50 mya), where we have much less information in the fossil record, and we have no direct measure of actual mutation rate, neither do we have full genomes on most these plants. All these things work together to make the molecular clock really bad in these cases. We are not troubled, because there are just to many unknowns to make a confident calculation.
In contrast, human/chimp evolution is about 6 million years, we have a much stronger fossil record, and we can and have directly measured mutation rates, and we have full genomes now too.
We found out a long time ago from studies like this that mutation rate varies a great deal. However, we did not know how to directly measure mutation rate till the last decade. That is what provides the independent verification.
@scd, you do not have to agree, but you have an opportunity to understand the thing you reject. Give that a shot.