Right, but to SpaceX it's just a commercially available off the shelf component. I'm sure they payed a pretty penny for it, but compared to the cost of building a custom building for integration, a penny is probably a pretty accurate estimate of the cost.
If you deal in Javascript all day a tube frame and keeping your bearings from corroding sounds hard but all this stuff is trivially solve-able. Big complex machines like this are the equivalent of a tall stack of software. No one part is complex. It's basically a given that it will work. Whether it scales is somewhat up in the air.
The most impressive thing about it is that it can be economically viable to operate while complying with the litany of relevant regulations across multiple jurisdictions. If you just need to stack rockets it's not hard to come up with a cheaper and better one trick pony that only lifts rockets. You see examples of such one trick ponies in every shipyard.
But SpaceX is constrained to some degree by engineering man hours so they were glad to just write a check for an expensive generalist piece of equipment. Cranes like this didn't exist in the 1960s so NASA built a bunch of traditional gantry cranes into a building. SpaceX can skip that sort of fiddling around and just write a check, so they do.
What you're saying doesn't really jive with reality though. SpaceX needs a specialist crane because they need to stabilize the rocket from the side, not the top, during launch. If it got hit by too much wind it would fall over otherwise.
