> NASA
Which proves my point. Look at the innovation that SpaceX has created with VTOL rockets. If it was left purely up to nationalized bodies this would have never happened (or maybe, it would have taken 50-100 years). The incentives and desire for economic efficiency just simply aren't there. Government wants flashy one-off wins that generate publicity and votes, if it wastes a few billion and doesn't scale there isn't any consternation. The entire mindset top to bottom is completely inconsistent with VTOL rockets.
> roads, railroads, airports, ports, GPS,
The argument for these being nationalized is that these are goods that are prone to natural monopoly and some of those are largely non-excludable in the case of roads.
None of those arguments apply to Amazon. eBay is a substitute for marketplace, Azure is a substitute for AWS, etc.
These are also much simpler than Amazon. Could a government have created AWS? Could a government create an iPhone? It's a rhetorical question because all know that the answer is no. Even if they could in terms of capability, they won't because the incentives aren't right. And if they tried, they'd do it incredibly inefficiently and be outcompeted in am embarrassing fashion by a private alternative (e.g. eBay), with the difference being subsidized by the taxpayer.
> nuclear weapons
In the case of nukes, the argument is that the government has a monopoly on force (which is also why the army should be nationalized).
And I already conceded that the army was well-run, although at a ridiculous level of cost (which is another byproduct of nationalization).
> weather satellites
Built by a private company.