No, China needs to do both. They need to catch up with existing technology and develop new technology at the same time. Developing new technology without understanding existing technology gets you in the place India is at this time. i.e. trying to spin gold from straw.
There has been talk that Moore's law is obsolete for decades. The reality is, there is a roadmap to keep it going for at least another decade, and it has been so for several decades. Once we get to the atomic level, then we will have to come up with something else, perhaps vertical integration, but it does not mean this technology is useless. Much like older nodes still get used today.
There was a huge delay with EUV and people thought we were going to be stuck forever at the same process node. Then someone came up with immersion lithography and gave the industry at least a decade more of shrinks in transistor size. Enough time that EUV is now working even if not at the scale it should be. The EUV light sources are still highly inefficient. Fact is the industry was working on EUV since at least the 1970s but basically only did basic research then and gave up. Just look it up. X-ray lithography.
"Extreme Ultra Violet" technology is in fact what used to be called "Soft X-ray" technology decades ago.
So as you see there is a path to continue die shrinks even beyond what EUV allows even if we have to got into "Hard X-ray" technology.
i.e. continue the path to decreasing wavelength. At least until we hit atomic limits.
I would disagree with you that Taiwan, Japan, and SK are only good at incremental improvements. One example is the violet laser which was basically invented in Japan along with modern green and blue lasers. This was the basis for Blu-ray technology and later for white LEDs for lighting applications. There was nothing "incremental" about it. People were trying to do it for decades and failed. Unless you think everything humans do is incremental. It was a titanic shift. Go look at research for batteries and graphene you'll also find a lot of Japanese researchers associated with it. South Korea also does a lot of technology advances but they aren't as well known in the hard sciences yet. Taiwan is limited by its population pool, but the business TSMC does hasn't been replicated elsewhere and not for lack of trying with vast resources sometimes.
