Setterfield's hypothesis is more of a stretch than you know.
Aside from the problem that the curves are arbitrary and it relies hugely on older - and less accurate - attempts to measure the speed of light it needs to make the following assumptions:
1) That the decay in the speed of light just happened to stop at the point where accurate measurements became available.
2) That other changes occurred so that just happen to mask any evidence of the change in speed. (For instance we should see a "slowing down effect" in observations of distant objects because the later the light is emitted, the slower it must go).
The "White Hole" hypothesis is supposed to be nearly as bad (it has been mauled by the Creationist astronomer Hugh Ross, for one).