It is possible to trace CJ transactions and has been for quite some time.
That might be true, but there is no denying that Wasabi coinjoins will be under much greater scrutiny than other coinjoins, since Wasabi are actively paying a blockchain analysis company to monitor their coinjoins and tell them if they have to censor any specific inputs.
Are you sure they are giving a list of inputs to a blockchain analysis company, as opposed to getting a list of addresses/inputs that should be blacklisted that they can compare proposed inputs to?
Even if it is the former, it would be unusual for a company to use data from their paying customers for other commercial purposes.
The broad use of such tools as Mercury wallet or coin swap may help to make blockchain's transparency a less reliable source of information for undesirable observers.
I completely agree, and I've said as much before - if everyone just started mixing their coins as their standard practice, then the concept of taint would disappear overnight and bitcoin would be completely fungible. Every centralized service would either have to accept any and all bitcoin, or go bankrupt. Bitcoin would be completely fungible, no one would end up with accounts being locked for arbitrary reasons, blockchain analysis companies would be useless, everyone would regain so much lost privacy, and the whole ecosystem would be far better off for it.
Not necessarily. With the exception of CM, researchers have been able to trace inputs to outputs from all major mixers, and have published their results. CM uses countermeasures that would prevent inputs from being linked to outputs using methods described in public research, however, in theory, the inputs may be linked to outputs (assuming of course CM is not actively keeping track of inputs/outputs pairs and/or is some kind of honeypot).