Thanks for your reply. The problem is not with the mining, that works using a modified cgminer with the golden-guy patches also on Bitcoin core 0.20 onwards. What I was worried about is that some additional changes might need to be done for taproot in cgminer in order to produce valid blocks.
All this taproot support signalling discussion made me uncertain, while on the other hand solo mining on testnet has worked just fine.
After having dug through the taproot proposal thread just now, I realized that since I am using the official Bitcoin core reference implementation for my solo-mining experiment, everything taproot related is already implemented and available. Didn't happen to think that many (if not most) pools are using a custom implementation and as such have to add necessary taproot code themselves - and signal their readiness once done and deployed. With this, all this signalling makes much more sense
So if that is the case, and someone can confirm this, the only question left open is whether the taproot signalling is supposed to be included in every new block after the activation of the soft fork on mainnet. On testnet this was not the case, but for mainnet this may be vital since the activation.