Sure it is. It is the best indicator we use.

Certificates, diploma, driver license, etc. are all based on that.

Scientific research is inherently a field where progress could be modeled as punctuated equilibria. Just because there are long periods where little progress is made on a hard problem does not mean that it will never be solved, or that the solution is perpetually "15 years away". But it does mean that attempting to linearly extrapolate progress over a period of time is likely to be incorrect.

"where little progress is made on a hard problem does not mean that it will never be solved"

Never said that. But if fusion scientists say since 50+ years, that they are allmost there, then it is a pretty good indicator, that there probably also won't be a working fusion plant in the next 15 year's.

That would be true if we had no other data to go on, but we do: the fusion triple product, which has to exceed a threshold for net power. The JT-60 tokamak in Japan reached a triple product in 1999, using D-D fuel, that would have barely exceeded breakeven if they'd been using D-T. Fifty years ago we were nowhere near that close.