Do we have the right incentives?

Taking a break from the usual "technical" tip, I'd like to ask a question about incentives in the system. Is there a strong incentive to communicate effectively? I ask this because it seems:

1. Many senior researchers seem to be too busy flying around giving talks to take the time to prepare a good talk. Is giving many lousy talks better than giving a few great ones?

2. Often researchers say that they don't have the 3 or 4 hours to improve a talk. This ignores the argument that each member of the audience may waste an hour listening to a talk they don't understand. The system doesn't seem to optimize the time of the entire community.

3. And perhaps most troubling, I've noticed that a few researchers really don't care if they are understood. Their primary goal is to record their talk on their CV.

I hope this is not the case, but what I see troubles me. Another indication of this is that many researchers use their full hour and don't leave time for questions. What is the point of giving a talk if there is no interaction? A journal article is probably easier to understand than a slide presentation.