I should have mentioned in my post that I learned of this paper from your blog.

And yes, exactly what you said. In my time working in special education, I saw plenty of cases where the client would try to communicate their needs, but the authority would go ahead and do what they wanted, because communications wasn't good enough for them.

And the idea of different types of intelligence leading to communication gaps has implications not only for AI research but SETI as well. I've often thought that the interest in intelligent extraterrestrial really boils down to "Beings enough like us that we can talk to them".