I see two questions here:
Is programming with core.async similar to programming in continuation-passing style?
Are the internals of core.async related to CPS?
The answer to 1. is "no", because there are no explicit continuations. In fact, this is kind of the point; one of the primary reasons for using core.async is to escape callback hell, which is basically CPS gone bad (callbacks being the continuations).
The simple answer to 2. is also "no", since the go
macro uses an SSA internal representation to produce a state machine. It is true, however, that SSA and CPS are related in that they are used for similar purposes (that is, as internal representations in compilers; they feel different on the surface and aren't completely equivalent semantically, but in implementing something like core.async you could in principle use either one).