He must have either seen it as important (like the gods being first in Menoikeus and the PDs) or fundamental to his physics.
I strongly think that Epicurus saw it as BOTH of those which you reference. It's a shame that his ideas on this have been dismissed they way they have. Once again, DeWitt is pretty good in taking Epicurus seriously, which apparently means considering the mind as a "suprasensory mechanism" able to receive and perceive these images directly without going through the eyes.
When we discussed this section of the podcast we really didn't dig into it too far, but even for those who reject the idea out of hand today, we need to keep in mind that this was important to Epicurus. It's even possible (and I am not advocating this at least at present) that this was tightly involved in the mechanism of "anticipations. "
Now that I write that, I do remember that when we were going over book 4 in the podcast that some of the panel thought that potentially the issue involved was that the receipt of images over time would imprint patterns that would be relevant to later thoughts, sort of like the citation in Diogenes Laertius, but instead of involving things we see, things that the "images" would imprint in the mind directly. It's very clear that Epicurus thought that these images were involved in our dream processes.
So at this point I would speculate that to the extent you are onto a good trail in regard to a teaching or therapy mechanism, the basis for it would be that Epicurus held that repeated exposure to images over time would imprint on the mind and affect later tendencies in thought processes.
It's a related point, but not the same point, to speculate that the receipt of these images is involved in anticipations, and that this is highly relevant (and may even "be") the third leg of the canon - the explanation for "intuition." I think we discussed in the podcast that the imprint of these images in the structure of the brain might be relevant to genetic transmission of dispositions across generations, since "something" is involved there, given that it strongly appears that Epicurus was not a fan of "blank slate" theory.
Regardless of whether either of these are fruitful to discuss or any particular users here think that they can be reconciled with modern science, it appears to me to be well founded that Epicurus himself thought something like this was true, so we should expect that he and his later school members attempted to follow these lines of thought to their logical conclusions.