E=mc^2. The proportion of the original mass converted to energy in fusion is higher than it is for fission.

The trouble with fusion is that you need really high temperatures and pressures to sustain it. You can easily sustain a fission reaction at a few hundred degrees C, or even lower than that if you wanted to (but the general goal is to boil water to run a steam turbine so >100°C is what you want).

Fusion requires millions of degrees, which would vaporize most things you might want to use as a container, so instead they use strong magnetic fields and then you spend a lot of energy maintaining the magnetic field. A bomb doesn't care about that because sustained operation isn't required and vaporizing everything in the vicinity is kind of the idea.

Moreover, efficiency isn't really the issue. The efficiency of fission is already absurd. Fission of a given mass of uranium generates the same amount of energy as burning 2.7 million times that amount coal -- by mass, it's even more by volume.

The main advantages of fusion are that hydrogen is a lot more common than uranium (though uranium is still pretty available) and that the byproduct of fusion is helium (a non-radioactive noble gas with general usefulness), whereas some of the fission byproducts are variously radioactive or have limited known commercial applications.

The history of fusion research for the past 70 years is trying to figure out a cost-effective way to trigger enough fusion to make net energy without a nuke. So far, we've got bupkis. (The magnetic containment approaches using high-temperature superconductors seem promising, but we won't know they actually work until we've built them.)

Thermonuclear bombs literally create lots of heat, and don’t tend to drive anything that produces work. It’s the opposite of efficient.

Just like lighting gasoline on fire and claiming that that’s more efficient than a combustion energy…

So, fission bombs are the black magic.