Re: "this approach relies neither on the cosmological model nor the cosmic-distance ladder"
"When we look at standard candle supernovas in all directions, we observe that the far away (and thus older) ones are moving away from us more slowly than we would expect if the expansion of the universe was constant. Conversely, the closer (newer) ones are moving away from us faster than expected if expansion was constant."
Hmmm... To do what you're saying we need to independently measure speed and distance of a star, and as I recall the whole Hubble thing, faster (more redshift) = further away, so we can use the redshift (ie velocity away from us) as a proxy for distance away. And the parallax + brightness 'ladder' method to measure distances surely to independently measure the distance to any star surely gets more inaccurate the further away you get.
It must be a pretty big difference between observed vs 'modelled for constant acceleration' to be able to say quite definitely that the expansion is accelerating. I am constantly in awe of how much we can learn across such vast distances with what are in effect quite puny instruments