i dont have pts downloaded so dont know for sure, but estimating at least 20%-40% depending on the toon

I'm on the console, so I'll have to wait to see. I've been told that as I lower-tier player I won't notice the U35 changes. So maybe the people who respond should mention their general skill level?

It might be useful to know if there actually is a difference.

Given that the 21mil dummy was buffed on the PTS, I had to parse on the 6mil dummy to ensure that a live vs. PTS comparison wasn't confounded. Parsing on the 6mil naturally meant that my overall DPS would be substantially lower than a fully buffed 21mil parse. However, the 6mil dummy is unadulterated by any buffs, so this at least allowed for an apples-to-apples comparison. (I deliberately chose the 6mil over the 3mil because the execute phase on the 3mil tends to be absurdly short--and therefore doesn't represent the potential of your execute damage. Not to mention, a 6mil parse gives literally twice as much data as a 3mil parse.)

I only tested my DK because I know it like the back of my hand (it's the class I use for most of my trifecta achievements). Despite knowing it so well, I still saw about a 20% drop on the PTS. I did 4 parses each on the live and PTS servers, and then averaged them to filter out the effect of crit RNG from individual parses. (My unbuffed crit on my DK is about 49%. At that level, under- or over-critting can have a large effect on the results of one parse. Hence why I did multiple parses and took an average.) My average on the live server parses was 41k; on the PTS it was 33k. The difference is technically 19.6%, but I rounded up to 20% for the sake of brevity when answering the poll.

Edited for typos.