All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=893533

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=7af5092c5f7283bbf91c97e43ceaa802eaf1603c4455c7e7710268142b6a7910


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

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

The metrics were changed first in https://bugs.chromium.org/p/chromium/issues/detail?id=889017

Then reverted back here due to one of these very innocently looking 3 CLs:
https://webrtc.googlesource.com/src/+/c411cdfbc31370aad1df2e4599338824feaf2cda
https://webrtc.googlesource.com/src/+/135aad0048df9151337bea3bd39b4c8d138f3d1d
https://webrtc.googlesource.com/src/+/637b0b5d38979a0610ef99e50925ea65b6bdf4b8

 Issue 893532  has been merged into this issue.

Most metrics did return to the previous level.

The only 3 not reverted completely are:
1) vp9 svc high dropped frames, stayed at 2-3 instead of 1-2
2) ksvc receiver time, which is down from ~55 do ~47 ms.
3) foremat cif vp8 encode time down from ~11 to ~10 ms.


Closing the bug as WontFix because changes are improvements or non-significant and all the CLs in range look completely unrelated.