"When somebody that ACTUALLY FLIES THE PLANE says..."it is hard to see what the PF is doing on an Airbus"...consider that this person has been there for real and is telling you his real life experience."
As I mentioned above, Bill Palmer who has nine years experience as an A330 Check Airman discussed this in his recent book. He also flew the AF447 scenario in multiple commercial simulators, although dual control conflict was only one factor in that case.
He explained the issue isn't simply whether the async sidestick design makes it difficult to see what the other pilot is doing. Rather it's what material difference does that make, and in what situation. He said: "While it is difficult to see the other pilot's sidestick position, there is rarely a reason to."
Also, dual control input is enunciated by a voice warning, plus a warning light on the glare shield. In the AF447 case, both pilots were also pressing their "takeover priority" button trying to lock each other out. Each time they did this an additional red warning light notified the locked out pilot his stick was taken over, plus an additional "priority left/right" voice warning. They obviously knew the other pilot was interfering, else they wouldn't be fighting over the controls -- repetitively. In a mechanically linked system it would be a "force fight", or maybe it would degenerate to a fist fight. That would be better in some cases but how representative is that?
On the opposing viewpoint, see this critical history of the AirBus async control design and various steps taken to improve it:
Pilots in the Loop? Airbus and the FBW Side Stick « Critical Uncertainties