Lanefan
Victoria Rules
Same here - except I've never used xp-for-treasure.We played published modules (as is) as well as a lot of homebrew stuff. We followed the treasure tables for monsters for loot, neither adding nor subtracting from the results. When I DMed AD&D (which was most of the time), it was always "let the dice fall as the may". If I rolled up a vorpal sword for a Giant Beaver a party of 3rd-level characters defeated, they got a vorpal sword. Sometimes that was a huge boon in xp (such an item was worth 10,000 XP after all!), as even divided by the 5 characters would be enough alone to be half the XP they need to level!

All three differ.Either way, as I have said repeatedly, awarded XP for treasure was typically maybe half (at best) of the total XP accumulated. Some times it was more, most times it was less. Now, it is subjective to the DM. If they want faster leveling, the award treasure at a higher ratio, using 1-1 as a baseline for an average encounter instead of something hard. I typically awarded 1 XP per 5 gp recovered for the average (think moderate in 5E) encounter.
As far as the per session comment, it has nothing to do with adventures. I am simply talking about the rate in playing time at which characters in 5E level compared to AD&D. Of course it varies from group to group. In 5E the suggested rate is a level per 2-3 sessions (4-hour), in AD&D we would be lucky to level in twice that time. In our CoS game we made level 6 in only 8 sessions (granted, we accomplished a lot), but that is pretty fast and was simply by the XP awarded for defeating what we encountered.
In the long run, if you disagree with my experiences and they differed from yours that is fine and just say so, but what someone else did as their analysis is immaterial to me and my experiences.
Your ratio differs from the analysis; yours is actually closer to what I'd expect. Mine differs from both in never having used xp-for-treasure. But for me the analysis is still relevant as were I to ever go to a xp-for-treasure system its in-depthness gives me a vague baseline as to what to expect.
As for comparing campaigns etc. within an edition or across editions, I've always found that using a by-adventure comparison works best as it filters out and ignores the variables of real time, session length/frequency, and so forth and thus gives a valid apples-to-apples measure.
And it's edition-agnostic: most D&D campaigns across all editions have been made up of reasonably discrete (and thus countable) adventures.