This CL: - makes it easier to understand the (nontrivial) metric interpretation - corrects the computation of BufferDelay to use 0 for absent delay - deletes metric MaxSkewShiftCount, unused since https://webrtc-review.googlesource.com/c/src/+/119701 - updates the unit test to directly test metric reporting Corresponding update to histograms.xml: https://crrev.com/c/3944909 Previous revert: https://webrtc-review.googlesource.com/c/src/+/279040 This CL is identical to the original, except: - the test is updated to spam fewer EXPECT_EQ failures on failure (EXPECT_EQs moved out of inner loop) - the test not resets metrics (metrics::Reset()) at the beginning, like other histogram tests Bug: webrtc:8671, chromium:1349051 Change-Id: Ie802e1f9d03a22ff7018f522a63b19e0b6eec2e8 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/279046 Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org> Commit-Queue: Sam Zackrisson <saza@webrtc.org> Cr-Commit-Position: refs/heads/main@{#38376}