It has worked that way for several decades, but as you say, an authoritarian state can change direction quickly. There are some indications that Xi's ideology is very similar to Mao's: total domination of everything by the guy at the top. Will it be a surprise when everyone starts carrying around a little red smartphone?

The interesting thing is that Xi seems smarter and wiser than Mao ever was. Well, in matters of war, Mao was really smart, and hopefully, we'll never have to see Xi tested that way because if we do, it'll probably be during a world war. But in terms of economic decisions, I'd say he absolutely tops Mao. The question is whether there's any situation where he wins in the long run because the responsibility of running China is basically being dealt a really bad hand; the issues are perhaps more complex and risky than corresponding issues when running the US.

I highly doubt Xi is smart in economics or whatever else, even compared to Mao, and he seems partly sincere in his belief in his methods, which raises some worries. One can argue whether Mao was really ideological or just supremely egotistical and corrupted by power. After all he was rather pragmatic in the beginning.

What is true though is that institutionally China is much smarter about economics and less ideological than in Mao's era, but that is predicated on retaining institutional norms and memory past Xi and his ostensibly tactical rollback. It could totally work out in a worse way.

Central decision makers generally cannot predict the outcomes of the decisions they make because the system is so complex. It's been proven time and time again that the distributed decisions of millions of individual actors ala a non-centralized economy results in a more efficient and non-catastrophic outcome.

Wherever recent history has gone awry is when a few powerful people think they know what's best for everyone else.

You have completely wrong image in your mind. We are not talking about fully centralized command economy.

Just like I said in other comment, Look at the post war economies in Europe, Japan, Korea etc. Wage controls, price controls, rationing etc. still a market economy and not a command economy.

An interesting line of thought. Some writers have suggested that the failure of centralization in prior eras was due to a poor state of information availability. If you consider why things like the Great Leap Forward continued till it became a catastrophe, one big reason was because politicians in locales made up good numbers for political reasons. As long as humans have been involved, this and other more technological barriers (as in the case of the USSR) have presented insurmountable challenge, but things are certainly moving in a direction that makes centralized decision making less impractical.

Note, this is still not a suggestion to do it, for humanist reasons, which I think is mostly what the Western (cultural) objection has always been. But from a practical perspective, centralization (even totalitarianism) has been accepted as the norm in various smaller and non-governmental contexts that have also involved highly complex decision-making, e.g. the corporate employer. People seem to accept this level of de-individualization, and a strong evidentiary case has yet to be put forth that full decentralization of that relationship works out better over all.