Rework latency/duration/endTime of Span

[Oberon00]

Split & rebased from #693. The only difference to #693 is that the latency/duration is not added to SpanData, only ReadableSpan.

In effect, this PR:

• Changes the end time of active spans to return zero instead of the current duration/latency if the span has not been ended yet.
• Add a public getter for a ReadableSpan's latency/duration.
• In contrast to [On Hold] Improve behavior of un-ended ReadableSpan #693, this does not add the latency property to SpanData (I think this was a bad idea anyway).

[codecov-io]

Codecov Report

Merging #737 into master will decrease coverage by <.01%.
The diff coverage is 100%.

@@             Coverage Diff              @@
##             master     #737      +/-   ##
============================================
- Coverage     78.86%   78.85%   -0.01%     
+ Complexity      754      753       -1     
============================================
  Files            94       94              
  Lines          2626     2625       -1     
  Branches        235      236       +1     
============================================
- Hits           2071     2070       -1     
  Misses          459      459              
  Partials         96       96

Impacted Files	Coverage Δ	Complexity Δ
...ntelemetry/sdk/trace/RecordEventsReadableSpan.java	89.41% <100%> (-0.07%)	48 <0> (-1)

---

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 f4fca3a...3256122. Read the comment docs.