All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=598776

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg8OyLvQkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg8JziqgoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg8MTSqAoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg8Nq6qwoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg8MSOuQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgiMDnpAkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg8NqhugoM


Bot(s) for this bug's original alert(s):

webrtc-linux-large-tests
webrtc-mac-large-tests
webrtc-win-large-tests

r12126 == https://codereview.webrtc.org/1833083002

I think that's the roll mentioned in  bug 597705 . The regressions now tracked in this bug don't seem as bad, and on closer inspection, most of those "regressions" still end up at values that are better than the ones noted before  bug 597705 . The one exception is this one:
https://chromeperf.appspot.com/report?sid=352d273b39d041c3f0ad6ce1a10dd15846f41a7fa1fff45df753c39ed91f0580&start_rev=11701&end_rev=12145

Anyway, marpan@, can you take another look? Overall, I would be ok with closing this as wontfix, but I trust your judgment more. :)

Oh, and I forgot to mention that I also associated a bunch of "improvement" graphs with the same alert.

Marking as won't fix. There is only one regression that didn't seen to recover from the roll last week, as mentioned in #2. And given the some of the improvements, won't try to fix.