Re: It's about government control
... but you could expect that for rocket science done by amateurs...
SpaceX's record on planning isn't very good. A long time ago their intention was to design a disposable booster that was so cheap to manufacture that it didn't matter that it was being thrown away every launch.
To this end they did do some quite clever things, including the original design of their rocket engine bells. These were made by forming two bells from sheet, pressing one of them to be crinkly, fitting one inside the other and welding / brazing them together. This made all the cooling channels for the bell in only a few operations; quick and a lot cheaper than brazing miles and miles of tubing into the shape of a bell.
Anyway, it turned out that they couldn't get the price down far enough that way. So disposability went, re-usability came in, hence their landing legs, etc.
And with Falcon Heavy, they seriously underestimated the cost and difficulty of strapping 3 boosters next to each other. Anyone with any knowledge at all of aerodynamics could see that the loadings were going to be horrible. The interference drag between the boosters must be epic.
There's also persistent talk of them not getting on top of a comprehensive quality control process. That led to their helium tank struts being rubbish (second stage explosion during first stage burn), and the helium tank failure on the launch pad (second stage explosion on the launch pad during static test; they'd not considered the full effect of super-cooled O2 on the carbon fibre tank). QC processes cost money, but no where near as much as losing a payload, or indeed a crew.
And when it comes to getting Falcon / Falcon Heavy man rated, meetings between NASA and SpaceX on this topic didn't go well; turns out you can't just claim it's reliable, you have to do all the paperwork to demonstrate that. NASA's manned flight loses were entirely down to things that were thought to be reliable or impossible turning out not to be so. So they're not willing to let anyone fly on a booster that just happens to have been reliable for a few launches in succession. That was some time ago, reported on El Reg somewhere, and perhaps SpaceX have been working towards getting their paperwork in order since, but it's extremely hard to do it retrospectively.
So yes, SpaceX's own history is no different to any other organisation trying to do things empirically. They've got caught out time and again.
Musk is, first and foremost, a showman, and on no account must his extravagances be permitted to hurt anyone. When it comes to manned flight, boring and expensive but done properly must override exciting / cult-ish and cheap but with doubtful paperwork.
NASA supposedly has all the know-how already. Or did all the competent engineers move to private industry by now?
The work being done on SLS is being done by private industry on behalf of NASA, and they do have the knowledge to make a man-rated booster. Re-using Shuttle engine and solid booster designs is an excellent way to achieve this.
NASA have never had large factories of their own for manufacturing rocket parts. They have the final assembly building, where it's all brought together.