**Why is this PR necessary?**
In clusters with frequent pod restarts, Prometheus may drop scrapes
due to duplicate timestamps reported by the exporter. Allowing
honorTimestamps to be disabled lets Prometheus use scrape time
instead, preventing those drops.

**What does it do?**
Adds an optional honorTimestamps field to the controller, proxy and
serviceMirror PodMonitors in the linkerd-control-plane chart. The
field is rendered only when set, preserving backward compatibility.

**References:**
N/A

**Notes:**
N/A

Signed-off-by: bezarsnba <bezarsnba@gmail.com>