[aggregator] Implement timedMetricList.FlushOffset

[linasm]

What this PR does / why we need it:
Changes timedMetricList to use fixed flush offset.

Special notes for your reviewer:
Dropping fixedOffsetFlushingMetricList interface and moving the FlushOffset() method to flushingMetricList (with ok flag) to avoid the error-prone type check in https://github.com/m3db/m3/pull/3292/files#diff-5a6baf266bd455b257e1a2f2ded077f3f2373a44628318748d2008d9ba59b9bcL284 which was masking the fact that timedMetricList was not exposing its flushOffset.

Old code:

	// Forwarded metric list has a fixed flush offset to ensure forwarded metrics are flushed
	// immediately after we have waited long enough, whereas a standard metric list has a flexible
	// flush offset to more evenly spread out the load during flushing.
	var (
		flushInterval = l.FlushInterval()
		flushOffset   time.Duration
	)
	if fl, ok := l.(fixedOffsetFlushingMetricList); ok {
		flushOffset = fl.FlushOffset()
	} else {
		flushOffset = mgr.computeFlushIntervalOffset(flushInterval)
	}

New code:

	flushInterval := l.FlushInterval()
	flushOffset, ok := l.FixedFlushOffset()
	if !ok {
		flushOffset = mgr.computeFlushIntervalOffset(flushInterval)
	}

Does this PR introduce a user-facing and/or backwards incompatible change?:
NONE

Does this PR require updating code package or user-facing documentation?:
NONE

[codecov]

Codecov Report

Merging #3292 (5875a94) into master (5875a94) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #3292   +/-   ##
=======================================
  Coverage    72.3%    72.3%           
=======================================
  Files        1098     1098           
  Lines      102020   102020           
=======================================
  Hits        73807    73807           
  Misses      23115    23115           
  Partials     5098     5098           

Flag	Coverage Δ
aggregator	77.0% <0.0%> (ø)
cluster	84.9% <0.0%> (ø)
collector	84.3% <0.0%> (ø)
dbnode	78.5% <0.0%> (ø)
m3em	74.4% <0.0%> (ø)
m3ninx	73.6% <0.0%> (ø)
metrics	19.8% <0.0%> (ø)
msg	74.3% <0.0%> (ø)
query	67.3% <0.0%> (ø)
x	80.5% <0.0%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5875a94...b98ac13. Read the comment docs.

[robskillington]

LGTM and definitely surprised this was not the case before