Without volumes, there is no privacy, that is why I am frontloading things that will create the volumes. What point to have perfect privacy on paper that in reality is trivial to brute force correlate due to small overall volumes?
I don't necessarily agree here. A big advantage of cryptonote is the lack of requirement for simultaneity.
There are certainly side channels that are possible if an observer knows something about your activity (for example that you are making a large purchase). But absent those side channels, private transactions can still occur even at a low frequency. I could send you 100 BTC worth of coins right now on XMR (BBR might be harder given the low total value) and while someone would know that a large transaction is occurring, they wouldn't know it was me sending it to you (as opposed me moving coins between cold storage, an exchange moving coins, etc.), nor would blockchain analysis ever be likely to determine this (unless, as I said earlier, they already knew that I was sending you coins).
All of these systems are susceptible to sybil attacks, which makes real activity somewhat valuable (in making sybil attacks more expensive). In the case of cryptonote the sybil attacker has to control a large portion of the entire output set, which makes the likely-fake activity on the BCN network even more suspicious. It is hard to argue that BBR is sybil attacked when hardly anyone is creating outputs!
Nobody disagrees with the premise and value of growing usage of a coin network, but coinjoin/coinshuffle-style systems make it more necessary to have a high level of real activity at nearly every moment in time, while cryptonote does not.