There is a lot of talk about what armor weights stam characters should run if the PTS changes go live. I was also wondering this, so I did some very basic tests to see if there was a clear winner. I tested on a stamsorc because that is my main stamina character. I should also mention I am looking at a PvE perspective. These tests were meant to be simple yet consistent rather than optimized parses, and even then the triplicate parses varied more than I would have liked. But again, I was aiming for something quick and dirty to get this discussion going.
I made full sets of medium Hundings Rage and light Diamond's Victory. All purple with gold weapons, all divines or robust with stamina and weapon damage glyphs. The weapons were front bar 2h sharp sword with stamina absorb enchant, back bar infused bow with weapon damage. I did four sets of parses in triplicate (repeated 3x to account for error) with different mundus and armor setups as shown in the results below. (Also, I used a 3 mil instead of 21 because penetration is relevant to the comparison and something I am interested in as a solo PvEer.)
Test 1: 5 medium 2 light Ritual mundus (to simulate no mundus boon). Avg dps on 3 mil: 22929
Test 2: 5 medium 2 light Lover mundus (increased pen). Avg dps on a 3 mil: 25627
Test 3: 5 light 2 medium Ritual (to simulate no mundus boon). Avg dps on a 3 mil: 23672
Test 4: 5 light 2 medium Shadow (increased crit damage). Avg dps on a 3 mil: 25655
In running these it looked to me like the 5 light/2 med setup had an edge over the 5 med/2 light, but it was also harder to sustain so lost some dps to HAs, especially on my final parse in test 4. If it weren't for that, the avg for test 4 would probably be a few hundred higher and thus match the comparison of test 3 to test 1 where light armor outperforms medium.
These results tell me that it will be worth considering switching to light armor on my stamina characters next patch, whether in a 5/2 configuration as tested, or in a 4/3 or 3/3/1 setup. The limiting factor will be set availability, but it's something to think about in the weeks before the patch hits live. Also, let me reiterate, these parses were done to compare armor passives and not achieve high numbers. I didn't run any monster pieces or mythics or bother with ults, just 7 Hundings 5 DV or vice versa with the skills I'm used to running. I'm also not that great at parsing
Please, if you have also done tests of medium vs light armor, I would love to see your results below.