This is in response to the KQ3 redux thread.
Ah, so is "saw" > "would" a valid construction? I would have assumed it were only valid if it were written "the truth you would have seen", but that's just going from my own instincts as an English speaker rather than my grammatical knowledge (which is sadly closer to nil than I'd like).
Consider this example lifted from a grammar site: "If I went to a friend's house for dinner, I
would take a bottle of wine or some flowers." We might add: "Back in those days..." because modern listeners/readers are more comfortable with these constructions when describing something which has changed in the present, but the point remains the same--it is possible to construct (thanks to the analytic/periphrastic nature of the English future tense) a form that conflates past and future, giving us a sort of "past-future" which stands between "back then" and "right now". It is in fact this conflation of the past and future which is responsible for the glossing of the construction as "conditional".
We wouldn't complain if the original read: "If you saw me, you would see the truth". Well, "If you saw me, the truth you would see" is grammatically equivalent to that and therefore logically equivalent (because semantics aren't separable from grammar). Now, should we exchange "when" for "if", what changes? We CAN still read "When you saw me,..." as a condition. The difference between when/if conditionals is that "if" has open modality, the condition is true subject only to its own internal modalities (the factuality of having seen or not having seen) whereas "when" has an additional modality applied, that of time.
If we say...
When you saw me, you saw the truth: the two events are contemporaneous (and unequivocally in the past).
When you see me, you see the truth: seeing the first person is EQUIVALENT to seeing the truth; again the events are contemporaneous.
When you see me, you will see the truth: the second person will see the truth (or come to understand the truth) some time after seeing the first; it is implied that seeing the first person is perhaps a necessary precursor to seeing the truth.
It is this sentence which has been cast into the past, so the sense is that upon seeing her, Graham has discovered something that precedes his seeing the truth.
We've done all of this without appealing to the progressive and perfect aspects. The perfect aspect just adds another level of recursion (our first level of recursion is the point of reference in the past tense... if we introduce perfects, then we're applying actions to points after the present and/or the past depending on where we stick our forms of to have

).
Also, a caesura is not technically a foot containing only one stressed syllable. I think the effect you're describing is:
Hrmm... yes, caesura is the pause within a line or the flow of speech... not sure where I first heard it classified as a single stressed syllable standing as a foot (though probably in high school as that was the last time I saw or did much thinking on prosody). It was probably a confusion with strong caesura which seems to be tied to the pause happening with the stressed syllable. In any case, my recommendation still remains to switch to a meter composed of only stressed syllables (spondee, molossus, dispondee... ask me again in a week... I'll have forgotten them

) to engage a sense of urgency or immediacy.