There are all sorts of approaches to fusion, and things such as type 2 superconductors were undiscovered 30 ago and uneconomic/unpractical 10 years ago. Timing control systems for magnetised target fusion were impossible but now are doable. Our understanding of plasma has been advancing a lot, simulations are good now, we can control plasmas much better. Chirped pulsed laser amplification is a thing now and really good at making high amplitude pulsed lasers for inertial approaches...
I could go on and on. This isn't the 90s anymore, and our technology is still rapidly advancing. What happens if we find more efficient/cheap/high power density thermocouples, or find a direct energy electrostatic power capture method?
Fusion's economic realities today may be overcome soon, we really do not know what we can do in even 20 years from now. The fundamental truth is that there is vast amounts of energy available in hydrogen, and all it takes is 100MK to ignite it.
And then you have the problem of having to stick sophisticated stuff in the hot zone where hands-on maintenance is impossible (compared to a fission reactor, where just the fuel and relatively simple hardware is in that zone.)
It's not clear why one should expect fusion to have good experience effects. Fission didn't, and the non-nuclear parts of fusion power plants will be mature technologies.
The engineering undesirability of DT fusion has been known for decades. All the recent excitement doesn't address any of the known showstoppers.
