Issue 36138: Improve documentation about converting datetime.timedelta to scalars

In a recent python-dev thread, there was some confusion about how to get something like `timedelta.total_microseconds()`. There is already an existing, supported idiom for this, which is that `timedelta` implements division:

    td = timedelta(hours=1)
    num_microseconds = td / timedelta(microseconds=1)

In this e-mail ( https://mail.python.org/pipermail/python-dev/2019-February/156351.html ), Nick Coghlan proposed that we update the documentation and there were no objections, quoting:

    * In the "Supported Operations" section of https://docs.python.org/3/library/datetime.html#timedelta-objects,
      change "Division (3) of t2 by t3." to "Division (3) of overall
      duration t2 by interval unit t3."
    * In the total_seconds() documentation, add a sentence "For interval units
      other than seconds, use the division form directly (e.g. `td / timedelta(microseconds=1)`)"


I am starting this issue to track that change.

I'll do this in a moment I'll have the PR

New changeset f40b4a0b6277b2779b9ded3736325489f2af93e4 by Nick Coghlan (Yasser A) in branch 'master':
bpo-36138: Clarify docs about converting datetime.timedelta to scalars. (GH-12137)
https://github.com/python/cpython/commit/f40b4a0b6277b2779b9ded3736325489f2af93e4

New changeset e213cd6325f2d6f116b6c3dce109cdf21f6263d5 by Miss Islington (bot) in branch '3.7':
bpo-36138: Clarify docs about converting datetime.timedelta to scalars. (GH-12137)
https://github.com/python/cpython/commit/e213cd6325f2d6f116b6c3dce109cdf21f6263d5

I've merged Yasser's PR (thank you!), but in the review discussion for that, we noticed something else: the `total_seconds()` documentation mentions the floating point dynamic resolution problem, but doesn't mention using `//` or `divmod()` as the fix for that, while the operator table doesn't mention the dynamic resolution problem at all.

So I haven't closed the issue yet, as fixing that feel in-scope for this issue, even though I didn't want to hold up the initial PR for it.