If being certain is being used to me "I know this as well as an human can know it" then that definition of certainty is practical and usable.
So I think that's what Epicurus was talking about when he was using his canon.
Yes.
So I think I would be concerned about granting to deductive logic or pure mathematics the status of "objective certainty" either.
Ah! Well-caught! One might say they are only "logically certain" -- in terms of the conclusions following necessarily from the premises. But not "objective" in terms of the real, empirical world (a deductive syllogism can be valid while leading to a result that is -- not logically, but empirically -- false). [Wittgenstein, for example, distinguished between what he called "logical space" and empirical space.]