So, update to this after talking to nikunjb@.

To ensure we can safely use this on the server-side we want the behavior to be like this:
- If tracking_only is true, if we were to show IPH, we still log the trigger to the DB, etc., but we return false to the caller.
- We only do this is tracking_only was the only failing "condition". It might not have to be a condition per se though, we could possibly just inspect it in the tracker itself.

Uploaded CL for this: https://chromium-review.googlesource.com/c/chromium/src/+/707636

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/cf591b593a17965b6fb2ea84413e909914f31675

commit cf591b593a17965b6fb2ea84413e909914f31675
Author: Tommy Nyquist <nyquist@chromium.org>
Date: Thu Oct 12 19:35:38 2017

Add support for tracking_only feature engagement configuration.

It was hard to compare experiment groups for in-product help rollouts,
since the tracking groups would never track their trigger state, since
the configuration was crafted such that the tracker would never return
true when asking to show in-product help.

This CL adds a new configuration field |tracking_only|, which when set
to true ensures that Tracker::ShouldTriggerHelpUI(...) will always
return false, but if all other conditions are met, it will still be
recorded as having been shown in the internal database and through UMA.

This is meant to be used by either local tests or for comparisons
between different experiment groups.

If one wants to later transition users with this flag set to `true` to
in fact display in-product help, they should use a different
`EventConfig::name` for the `event_trigger` configuration than the
non-tracking configuration.

BUG= 761084 

Change-Id: I8aefb85465eeb8ac4be24c5730c71ce2d9d2b961
Reviewed-on: https://chromium-review.googlesource.com/707636
Commit-Queue: Tommy Nyquist <nyquist@chromium.org>
Reviewed-by: Ilya Sherman <isherman@chromium.org>
Reviewed-by: David Trainor <dtrainor@chromium.org>
Cr-Commit-Position: refs/heads/master@{#508379}
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/components/feature_engagement/README.md
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/components/feature_engagement/internal/chrome_variations_configuration.cc
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/components/feature_engagement/internal/chrome_variations_configuration_unittest.cc
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/components/feature_engagement/internal/configuration.cc
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/components/feature_engagement/internal/configuration.h
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/components/feature_engagement/internal/stats.cc
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/components/feature_engagement/internal/stats.h
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/components/feature_engagement/internal/tracker_impl.cc
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/components/feature_engagement/internal/tracker_impl_unittest.cc
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/tools/metrics/actions/actions.xml
[modify] https://crrev.com/cf591b593a17965b6fb2ea84413e909914f31675/tools/metrics/histograms/enums.xml

From the time this landed, I believe this in fact landed for M-63.