How large? Because that's actually good because it could show that there's enough user-demand for different implementations of CoinJoin apps.
Not nearly as large as the group that has already upgraded to WabiSabi. ~3x-4x more new volume is coinjoined using the WabiSabi coinjoin protocol compared to the traceable Whirlpool coinjoin protocol, and Whirlpool further splits that tiny amount of liquidity remaining into 4 different pools, fracturing your anonymity set even further.
No, I meant in general. How much of the total volume of coins used in mixing went to centralized mixers/tumblers and how much of them went to CoinJoins - Wasabi's implementation, Samourai's, and include JoinMarket/other major implementations. If total volume is going up + the ratio of centralized mixers and CoinJoin is also going to the side of CoinJoin, then there's probably enough market growth and incentive for centralized mixers to slowly transition their business models to run CoinJoin coordinators.