First failure to focus on is the current failure mode:
Both stories for this benchmark are timing out:
TimeoutException: Timed out while waiting 60s for IsJavaScriptExpressionTrue.
Console output:

Starting in this run:
https://build.chromium.org/p/chromium.perf/builders/Mac%2010.11%20Perf/builds/1185

Before this run only the Doc story was failing with the following error:
ReplayNotStartedError: Web Page Replay failed to start. Log output:
2017/09/16 02:56:26 Loading cert from /b/s/w/ir/third_party/catapult/web_page_replay_go/wpr_cert.pem
2017/09/16 02:56:26 Loading key from /b/s/w/ir/third_party/catapult/web_page_replay_go/wpr_key.pem
2017/09/16 02:56:26 Loading script from /b/s/w/ir/third_party/catapult/web_page_replay_go/deterministic.js
Error openining archive file: gunzip failed: gzip: invalid header

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8967911079667112032

In:
https://build.chromium.org/p/chromium.perf/builders/Mac%2010.11%20Perf/builds/1181

The failure is:
Unexpected Failures:
* v8.detached_context_age_in_gc/https://www.google.com/#hl=en&q=barack+obama

ReplayNotStartedError: Web Page Replay failed to start. Log output:
2017/09/15 11:22:02 Loading cert from /b/s/w/ir/third_party/catapult/web_page_replay_go/wpr_cert.pem
2017/09/15 11:22:02 Loading key from /b/s/w/ir/third_party/catapult/web_page_replay_go/wpr_key.pem
2017/09/15 11:22:02 Loading script from /b/s/w/ir/third_party/catapult/web_page_replay_go/deterministic.js
Error openining archive file: gunzip failed: gzip: invalid header

In

The failure is:
Unexpected Failures:
* v8.detached_context_age_in_gc/Docs (1 open document tab)

ReplayNotStartedError: Web Page Replay failed to start. Log output:
2017/09/15 16:51:37 Loading cert from /b/s/w/ir/third_party/catapult/web_page_replay_go/wpr_cert.pem
2017/09/15 16:51:37 Loading key from /b/s/w/ir/third_party/catapult/web_page_replay_go/wpr_key.pem
2017/09/15 16:51:37 Loading script from /b/s/w/ir/third_party/catapult/web_page_replay_go/deterministic.js
Error openining archive file: gunzip failed: gzip: invalid header




1182 contains this cl:
https://chromium-review.googlesource.com/668685

That switches the wprgo archive used for the facebook story from top_25_011.wprgo back to top_25_009.wprgo

After the newest failure mode is [fixed/reverted/understood] (commented on in comment #1 and being bisected in comment #2), I would suggest we re-record any story that has its wprgo archive set to top_25_011.wprgo.

Yup, rerecod the benchmark is the right course of action.

=== Auto-CCing suspected CL author vmiura@chromium.org ===

Hi vmiura@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


=== BISECT JOB RESULTS ===
Test failure found with culprit

Suspected Commit
  Author : Victor Miura
  Commit : 26f31fe066aa26cbbd9cd2a1ea2cf32d4b52edfc
  Date   : Sat Sep 16 06:43:40 2017
  Subject: telemetry: Re-record top_25 google sites, and fix login flow.

Bisect Details
  Configuration: mac_10_11_perf_bisect
  Benchmark    : v8.detached_context_age_in_gc
  Metric       : V8_DetachedContextAgeInGC/https___www.google.com__hl_en_q_barack+obama

Revision             Exit Code      N
chromium@502495      0 +- N/A       2      good
chromium@502496      1 +- N/A       2      bad       <--
chromium@502497      1 +- N/A       2      bad
chromium@502498      1 +- N/A       2      bad
chromium@502501      1 +- N/A       2      bad
chromium@502506      1 +- N/A       2      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=https...www.google.com..hl.en.q.barack.obama v8.detached_context_age_in_gc

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8967911079667112032


For feedback, file a bug with component Speed>Bisection

After running some tests locally I think re-recording and then trying to debug any further problems is the correct route to go. 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d49a98a3cacdab8fbb2701ec67e44cfa9bd4cdf5

commit d49a98a3cacdab8fbb2701ec67e44cfa9bd4cdf5
Author: rnephew <rnephew@chromium.org>
Date: Thu Sep 21 00:44:02 2017

[Telemetry] Re-record top_25 stories used in v8.detached_context_age_in_gc.

Bug:  767114 
Change-Id: I1248fc3865e230c1ad52d5c5013b5a4501a93a4a
Reviewed-on: https://chromium-review.googlesource.com/676197
Reviewed-by: Ned Nguyen <nednguyen@google.com>
Commit-Queue: rnephew <rnephew@chromium.org>
Cr-Commit-Position: refs/heads/master@{#503300}
[modify] https://crrev.com/d49a98a3cacdab8fbb2701ec67e44cfa9bd4cdf5/tools/perf/page_sets/data/top_25.json
[add] https://crrev.com/d49a98a3cacdab8fbb2701ec67e44cfa9bd4cdf5/tools/perf/page_sets/data/top_25_012.wprgo.sha1

It has started to pass on most of the bots. There are a few that haven't completed a run yet however. When those finish and the benchmark passes; I'll close the bug out.

The v8.detached_context_age_in_gc benchmark is removed:
https://chromium-review.googlesource.com/c/chromium/src/+/980875