The US trade deficit with China declined from $418 billion in 2018 when the tariffs went into effect to $310 billion in 2020, so the tariffs have succeeded in reducing the trade deficit with China somewhat.

Obviously, as LC pointed out and Michael Pettis tracks closely, the overall US trade deficit with all countries has not declined (and has even slightly increased), so that means the tariffs just shifted production from China to other countries (as you noted, of course sourcing apparel had already begun to shift as Chinese wages increased--a friend has been heavily involved in this for many many years--but the data suggests the tariffs also caused a lot of movement out of China in addition to the apparel sourcing trend).

The Pettis/demand-side framework applauds the tariffs only in this way--since Chinese wages/the household share of GDP is so low in China, by definition they are able to buy less from us than we from them (because consumption is the main driver of imports, and their consumption is radically lower than ours--just 36% of their GDP vs 68% of the US's), so the structure of income distribution in their economy guarantees they will run a trade surplus with us.

I say guarantees, of course in classic macro theory, their currency should appreciate, but again as LC pointed out that Pettis writes a lot about, China then buys dollars in various ways, which is another way of saying when we buy their stuff, a relatively low portion goes to households who mostly spend the money and a relatively high portion goes to business/government who mostly save, so China has all this extra cash, a lot of which is used to buy dollars to keep the currency low.

That's the vicious cycle we are in, Chinese households benefit from employment since they supply some US consumers, but are badly hurt by low wages/the low household share of GDP in China. Chinese businesses, business owners, and SOE shareholders greatly benefit since such a high % of economic activity ends up in their hands, and Wall Street in the US benefits from conducting the transactions where China invests a lot of those business/SOE profits abroad. US workers are hurt because a lot of their employment goes to China.

The solution would be for China to rebalance its domestic income distribution so that Chinese households receive a radically higher share of national income (say, up from their current 50% to the US's 75%). That would reduce Chinese exports due to higher wages and a stronger currency (since the gov't would no longer have such a huge pile of money with which to buy dollars), but dramatically increase the size of their domestic market, so workers who previously worked for export companies could then work for companies supplying the newly much larger domestic economy.

It would also benefit US workers who would be able to export more to China, and who would not be competing with products from China made with suppressed wages and a depressed currency. So, workers in both China and the US would benefit from China rebalancing, and the financial sector in the US and the owners of private businesses and shareholders in SOE's in China would be hurt since their share of national income in China would decrease from 50% currently to something much lower.

This framework always has seemed right to me, the classic laissez-faire interpretations seem overly ideological and rigid and cannot explain why this view is wrong, so it's the view I've adopted. Hope this summary was helpful to you.