Re-generate library using speech/synth.py

[danoscarmike]

Incorporate new changes to Speech v1 per latest commit to googleapis.

[danoscarmike]

@theacodes seems to have overwritten the README pretty heavily -- reverting from Beta to Alpha -- that's not expected is it?

[theacodes]

@danoscarmike probably not. We can write a quick replace in synth.py to fix that, something like:

s.replace('README.rst', '`Alpha`', '`Beta`')

[danoscarmike]

Done and regenerated. PTAL.

[theacodes]

Thanks for doing this, @danoscarmike!

[tseaver]

@theacodes We likely want to work out how synth works for docs for manual veneers: having it step on README.rst, docs/index.rst, etc. is probably not what we want in those packages.

[theacodes]

@tseaver the solution there is to selectively copy files and use replacements as needed.