Helion Energy has a contract to deliver Microsoft a fusion reactor by 2028. It's only four years away so we don't need to speculate about it. Helion will either deliver or they won't:
https://www.reuters.com/technology/microsoft-buy-power-nucle...
https://www.youtube.com/watch?v=_bDXXWQxK38
MIT's Commonwealth Fusion Systems has started construction of their small scale SPARC reactor. Once that has proven itself they will build a larger scale ARC reactor. They should get to first plasma in 2025:
https://www.youtube.com/watch?v=w3Giq6NuPYs
https://energy.mit.edu/news/mit-designed-project-achieves-ma...
https://www.technologyreview.com/2024/04/23/1090425/mits-sup...
If it takes 7 years to build something that is known and has been done a couple of hundred times before, I think it's safe to say it will take more than 4 years to build something of similar complexity that has never been done before.
https://en.wikipedia.org/wiki/Helion_Energy#Development_hist...
The prototype currently under construction is Polaris. It's 25% larger than the previous prototype and they expect it to achieve net electricity production:
This is ultimately a capital game and without enough capital researchers and engineers have to go with the safest bets lest they lose funding. So you get ITER which will almost certainly work but has taken forever to build because it's a gargantuan feat of overengineering in all the ways that matter to make sure it works.
And for what it's worth, ITER's new baseline schedule will be announced this upcoming Wednesday (July 3). Likely first plasma will be in the next 2 or 3 years with first fusion operation some time between 2035 and 2040.
i.e. we know a well defined lower bound on the conditions required for energy positive fusion.
These bounds are more or less what we expected so sustainable energy positive fusion is essentially guaranteed to be feasible, it's just ungodly expensive until we can learn how to reduce the lower bounds for fusion or improve our ability to achieve those conditions.
So ITER will work. There really isn't a discussion on whether ITER will work or not. Like technically there's a small fraction of a percent chance it won't work for a number or reasons but even then it's less likely to be "it will never work" and more "we need to make changes and it'll work eventually".
This is actually part of the reason for the updated baseline schedule. It's apparently going to be far more aggressive now that there are better assurances that the existing model is in line with experimental evidence.
ITER won't be a power plant of course. It's a research facility first and foremost but it will produce power at a 10 to 1 ratio relative to the input which should be a net 450GW of power before you factor in the secondary equipment.
Containment and chilling will require quite a bit of energy of course but all together the energy usage for the reactor facilities should still be less than the produced power.
But yes ITER isn't designed to produce power out to the grid. It's just designed to work sustainably and any power it does produce will almost certainly just be sunk into a resistive load on site. It's a research project to build a viable nuclear power plant first and foremost.
----
Then provided ITER doesn't outright fail, the DEMO reactor will be built and that facility will be optimized for power generation and will be built to the requirements to efficiently produce energy (vs ITER which is significantly overbuilt "to produce power at Q=10 so help me god"). The DEMO project will of course actually hook up to the grid with the intent of being a sustained base load power generator.