NASA is very good at engineering solutions to problems.

In the context of SLS, the real obstacle wasn’t technological, but political. The predecessor to SLS, the Constellation program, which was a reasonable plan, was cancelled due to political horse-trading. Thus, NASA engineers came up with a program that would be politically un-killable due to how many different states and congressional districts the plan specifically involved.

To be more clear: NASA could have primarily engineered it for reusability, efficiency, budget, or any number of other factors, but they recognized flimsy budgetary commitments as the primary threat to the program (since that did, in fact, kill the predecessor program), and so they designed the SLS program to address that risk element first and foremost. In that context, it makes a ton of sense.

So yeah, it’s not reusable, or efficient, and it costs an absurd amount of money. But that’s because politicians keep screwing with these programs and often setting the projects (either intentionally or unintentionally) up to fail. And when you stick a bunch of engineers in a situation where they have to be responsive to non-engineers, and the engineers have little (if any) strategic input into the process, the engineers will design the project such that it has the best chance of success under the imposed constraints. For NASA, imperfect thing that exists is better than a perfect thing that won’t get past the congressional committee review, or one that will get axed a few years down the line.