This is good! But I had to read it twice and make little pictures to convince myself it was right.
I'm going to see if I can make it clearer.
We assume the ring radius is 1 unit (which is 1 light year).
We assume light was travelling at 3 times c at the time of the super nova.
We assume light slows down smoothly over the travel time to earth (I don't think it matters)
In the constant c scenario 1 year after the supernova blew we have the direct beam to earth 1 light-year along the way to earth and the detour beam to be just at the ring and not started to earth at all.
In the variable c scenario 1 year after the supernova blew we have the direct beam to earth as 3 light years to earth and the detour beam as having covered the distance to the ring AND cover 2 lightyears on the way to earth.
Thus in either case the detour beam is 1 unit (1 of our current lightyears) behind the direct beam. It doesn't matter how c has varied.
When the two beams finally reach earth the direct beam is 1 unit ahead of the other. If c has slowed down to it's official value we measure today the detour beam will take another year to arrive (if the unit was one lightyear of current c). We will then measure the ring radius to be 1 light year (at current c) no matter how light speed varied at the time.
I don't know if that helps or not. I did me good anyway