Thanks Eugenios. I think this is really the topic of Philodemus' "On Methods of Interfence" and does not in any way contradict your statement "one would think Epicurus would want to get the most accurate view of the universe." I completely agree with you.
The issue as I see it is sort of a preliminary rule of evidence, like a judge ruling on what comes into court. As you probably know there are elaborate rules of evidence about things such as hearsay that have evolved over time so that certain kinds of out-of-court statements are allowed in fully, or allowed in for limited purposes, or are kept out entirely.
And there is also a court parallel in regard to expert testimony as a whole, with very elaborate rules about when and how experts are allowed to testify, so that the expert does not do things such as "invade the province of a jury."
Another consideration that these rules of court apply to is to prevent "speculation" by the jury, in order to ensure that all decisions of a jury are based on evidence, and not left to simple speculation without evidence on which to ground it.
As I see it, it will never be possible to develop exact rules of "what comes in" and "what doesn't come in" and so in court, judges have to examine the facts of each case and take testimony and hear both sides and then evaluate whether to let the jury hear the testimony at all, which serves as a sort of "gatekeeper" function.
That's where I think Epicurus was going. He was saying that we all need "rules of evidence" so as to decide what kind of evidence is open to any kind of consideration at all, and what kinds of "evidence" should be thrown out of court and not even considered. Issues of claims of divine revelation would probably fit the type to throw out entirely - unless there is some other proof of the communication being claimed, someone saying "God told me to" is not even going to be listened to as evidence, other than perhaps evidence of insanity.
It will take a long thread and discussion to go through all of the examples, but as I understand part of the crux of the problem was that unlike our mathematicians and geometers of today, those "scientists" of that period were using math and "science" to argue that the supposed "order" they were finding was proof that the world was governed by divine commandment. They were arguing that the alleged hugeness of the stars was evidence of their divinity, and that the earth's place in the center of the universe was proof that it was specially ordained by god.
As such, Epicurus might not have been concerned with their calculations as such, but he was concerned with OVERREACH of their calculations to support theories that were not in fact supported by their contentions. I think that excerpt from Lucian maybe illustrates this as much as anything other.
I realize so far that we've barely introduced the topic, much less made any headway in discussing it, but it's my understanding that this probably lies behind the ejection of "reason" from the canon, and many other attitudes by Epicurus. Again, the best text I have found so far to discuss this is "On Methods of Inference" and the DeLacy commentary at the end of this edition.