Clean up handling of Handles

[oremanj]

• Use regular asyncio.Handle and asyncio.TimerHandle in most cases, reserving our custom extensions for the few places that actually need the extra features
• Name our extended handles in a way that clarifies what features they provide: ScopedHandle and AsyncHandle (also, two types rather than one type with a flag that changes its behavior and valid methods)
• Remove the logic for cancelling a Task when a handle for its next step is cancelled; this was never invoked in the test suite and it's not clear to me what it's intended to help with. (@smurfix can you provide any historical perspective on the rationale for this logic? It was added in 21f4096e "Fix hang-on-close tests")
• Move the logic for propagating the result of an async handle to a Future into the handle code (allows us to remove the "is_sync=None" special case for invoking the callback with its handle as argument) and robustify it; it should correctly handle MultiErrors now
• Run async handles' callbacks (trio_as_aio tasks) in a copy of the contextvars context that was current when creating them. The previous code appeared to do this, but didn't actually.

[codecov[bot]]

Codecov Report

Merging #76 into master will increase coverage by 1.84%. The diff coverage is 97.87%.

@@            Coverage Diff             @@
##           master      #76      +/-   ##
==========================================
+ Coverage   78.05%   79.90%   +1.84%     
==========================================
  Files          11       11              
  Lines        1285     1229      -56     
  Branches      180      171       -9     
==========================================
- Hits         1003      982      -21     
+ Misses        204      173      -31     
+ Partials       78       74       -4     

Impacted Files	Coverage Δ
trio_asyncio/_base.py	83.96% <94.28%> (+2.14%)	:arrow_up:
trio_asyncio/_async.py	79.41% <100.00%> (+5.88%)	:arrow_up:
trio_asyncio/_handles.py	94.93% <100.00%> (+17.57%)	:arrow_up:
trio_asyncio/_sync.py	70.99% <100.00%> (-0.22%)	:arrow_down: