My experience has been that experts in most fields tend to shy away from using precise, professional language when speaking to a layperson. I think this tendency indirectly results in the sort of dialogs you're describing.
The best we can nurture is 'good' analogy, that still somewhat valid even if the analogy is expanded further
p/s: If only we have a list of Feynman-like experts in most of the fields :D