Part of it is the sophistication. Take the Tomahawk: assumed range of ~1000 miles , estimated accuracy of 30 feet. Can launch from above or below water. Etc.

The other part is the limited production runs. Until last month, the DoD was generally purchasing ~100 of these annually. There's no scale economy in making these, so those 100 missiles need to support the entire production & R&D of the product.

Check out Joe Barnard's youtube channel BPS.Space where he's documenting his development of "high power" (hobby) rockets. Those are relatively small rockets still but nonetheless he's getting into performance regimes where the engineering starts to be tricky and the details really matter. The more extreme your rocket gets, the difficulty really ramps up quick.

It’s not a complete explanation, but I was awed by the precision of the shower screens used in modern rocket engines. In the 60s it might have sufficed to just spray fuel into the combustion chamber using some nozzles, but now we have highly precise matrices of micro-perforations that maximize combustion.

Also if you want to harden the rocket against EMP attacks you need an inertial guidance system, and those things also demand extreme precision.

Fireworks rocket do not cost as much. But if you want high precision and high speed, that simply is expensive. Also the area is of course restricted making it more expensive as most states do not want DIY rockets everywhere.