You can look up latency and it has a more general meaning. We're talking about a time delay. In this context we're not concerned about the time it takes to tranfer a single TCP packet, we're concerned about the time it takes to transfer a complete frame of video (under the assumption that a complete frame is required before the receiving end can display it). So bandwidth plays a role, and if you're sending larger frames then the delay from source output to destination display will be larger. This is latency.
Of course it does, but we're not talking 10Mb/s networks here. People stream 1080p content all the time, gigE can handle that no problem. The only latency concern would be the encoding step, not transferring the data.
