I also think that echolocation has potential to be equal or better to eyesight! Bats and dolphins do great with it, and I think that it's easy to think of echolocation as being a _visual_ stimuli, even though it's not lightwave-based.
I think the biggest hurdle for an underwater sentient species trying to become more advanced is the difficulties with tool-building. Simple stone tools could be made, but it becomes a lot harder to smelt iron and start doing metal-working.