Reland "Enable 'Browsing data volume counters on Desktop' by default."

Reland "Enable 'Browsing data volume counters on Desktop' by default."

This reverts commit 93f190df6d1e0435e535a4336e1fe9035a5c9eb7.

The original CL is https://codereview.chromium.org/1413373012/
The revert CL is https://codereview.chromium.org/1423373006/

The problem which caused the revert was the assumption that ProfileSyncService exists, which should be solved by the CL https://codereview.chromium.org/1413373012/ that landed in the meantime. The trybots did not catch it, as this issue was introduced after running git cl try on the original CL but before it landed.

BUG=553640
TBR=stevenjb@chromium.org,isherman@chromium.org,dbeam@chromium.org,petermayo@chromium.org,robliao@chromium.org
NOTREECHECKS=True
NOTRY=True

Committed: https://chromium.googlesource.com/chromium/src/+/ca05a1211ac818518364504b0fd73d5c6419aca2

[commit-bot: I haz the power, 2015-11-10 22:03:21]

Patchset 1 (id:??) landed as
https://crrev.com/ca05a1211ac818518364504b0fd73d5c6419aca2
Cr-Commit-Position: refs/heads/master@{#358922}

[msramek, 2015-11-10 22:03:52]

Committed patchset #1 (id:1) manually as
ca05a1211ac818518364504b0fd73d5c6419aca2 (presubmit successful).

[Peter Mayo, 2015-11-10 22:21:12]

On 2015/11/10 22:03:52, msramek wrote:
> Committed patchset #1 (id:1) manually as
> ca05a1211ac818518364504b0fd73d5c6419aca2 (presubmit successful).

It seems like bad form in general to leave "NOTREECHECKS" and "NOTRY" on reland
CLs.

They broke unexpectedly the first time, and in this case something which affects
success on landing has landed since.

I appreciate that their were a number of try jobs launched successfully, but if
we were that good at predicting interactions we wouldn't need the CQ in the
first place.

Letting the CQ do it's job may slow the change, but should improve the overall
velocity of the project, which is generally more important.

FWIW,
Peter.