Starlink is in a completely different class of "tons of satellites" as other platforms simply due to the SpaceX launch vehicles.
China has a total of 600-ish satellites together for all purposes. Starlink has 3500-ish right now, and the already licensed plans are for 12000 by 2026. When(if!) Starship starts launching satellites, a single launch can put in orbit more military satellites than China has ever launched, and it does not seem have a comparable capability in the near future.
You underestimate how incredibly difficult the work that SpaceX is doing is. No spaceflight oraganization in the world is close to having the capabilities that SpaceX has right now (Falcon 9, Falcon Heavy). Meanwhile Starship is almost ready.
That's why you do industrial espionage though; take advantage of all the R&D done by someone else, so most of the pitfalls are overcome and a lot of the testing is done.
Like knowing what kinds of alloys turned out not to work and why, etc. So you don't have to repeat all that.
Honestly this espionage thing is overrated. Plenty of employees in the US take their knowledge to (domestic) competitors and find it quite hard to just clone a business. The Russians and US were stealing each other's plans during the space race but this often led to misunderstood and suboptimal designs. See Buran
looking at the launch numbers from 2022 you can see that china is fairly close to having the capabilities of the us. Falcon heavy is still several years away from being useful in any capacity.
SpaceX is launching more than twoce the payload to orbit than the rest of the world combined though... without starship.
Also how is Falcon Heavy years away from being useful?
If you don't have a large enough LEO constellation you will have sporadic and predictable coverage. This is especially true if you want global coverage to monitor a military like the USA with a global presence.
1. Wouldn't these balloons provide even more sporadic coverage? They seem to be detectable.
2. These gaps would be measured in minutes, and are zero when you considered the combined coverage of all none-nato countries. Anything that we don't want seen is done under cover.
2. If you can scramble a jet within mins, that may be all you need. If you consider that LEO internet projects like SpaceX and Kuiper are targeting 5-10k sats to cover the earth i don't think it's a given that "all non-nato countries" have enough to provide constant coverage, especially if you go back 5-10 years. That probably won't be true in another 5-10 years.
1. Some reports say that they've only been detected recently; referencing multiple undetected occurrences during the Trump presidency; o they've been working for X years already. If they're undetected, they can view into gaps referenced above. If this program has been going on for 5-10 years undetected then it still could have been of value to China/Russia/Whoever owns them.