One of the things that I think I noticed from the press conference, is that funding is going to be the bare minimum to meet some goal for a design they select.

This seems like a gross mistake.

If we are going to avert a climate catastrophe we will need TW of power to "unburn" the carbon we put into the environment (ocean and atmosphere). Instead of barely hitting this target, we should over deliver since we are running out of wall-clock time.

Every project that meets a bar for feasibility, organizational/operational capabilities (if they dont have it, either fix it, or transfer design to capable team) should be given funding (50-100M). We should be dropping BILLIONS on this, if we can drop 50B+ on semiconductors we can do the same for fusion.

Dump trillions of dollars into fusion energy today and it will still be decades before the first fusion power plant is connected to the grid. You'd be better off funding the construction of fission power plants. Those are very expensive and take years to build, but they're still a hell of a lot cheaper and faster than funding fusion to the degree you're suggesting.

Each dollar diverted to chase nuke wills-o'-th'-wisp brings climate catastrophe nearer.

Money is fungible. Dropping $billions on this means not dropping those $billions on something that works already, works fantastically well, and would work even better with more money. We already know how to prevent (more) climate catastrophe. We just need to do more of it.

So drop it on fission. There. Settled.

Fission means, in practice, paying enough for coal generation, over the decade, to have built enough solar to displace the nuke; and paying many times that, on top, to build the nuke.

So, no. Each dollar diverted from building out solar to mining coal or fooling with nukes brings existential catastrophe nearer.

Beginner question, why is it hard(not possible) to calculate the gain without actually building a reactor?

Surely if you have a non-satisfactory yield you shouldn't be building it. And that should be possible to estimate before we spend billions on it?

Simulations and estimations about processes in the physical world always leave room for surprises when one is doing things that haven't been done before.