The speedoflight-stop-speedoflight (whomever said that upthread–thanks!) makes me think of the (a) conceptual problem for me and partly OP, I believe, in the question and answer byplay:
The word “light” is taken two ways, in the absence of further elucidation (heh). We’re told, and answered here, “light” slows down, because the “speedoflight-stop” nexus arrives as a bunch like that.
A complicating pre image has two components that “a” photon–is “that” photon and that perhaps more importantly, any thing that merits being “a” thing, if it undergoes the stop-"backup"up and we buy into that, just like we’ve bought into a reality of speed-of-light is a “real” time, we expect each “stop” to be a real time.
Which it is, but not at the energy-interaction moment within that traveling nexus which is accountable (i.e. countable) as such only in retrospect when we get hit by it). “That” photon that got “stopped” had no time to ramp-up, unstop instantaneously, whatever image floats your boat.
That photon is history and to the next in line in creation/energyform there was no “stop.”
ETA: so the image of “it took x years” for that photon to travel through the sun’s layers to reach us" is -fundamentally different (except for the statistical inferences) than the image that, say, “a molecule of Napoleon’s fart travelled x miles in the atmosphere to reach someon’e nose.” Which has been OP’d, BTW.