AWS API gateway timeout makes config timeout irrelevant

Hello,

The AWS API gateway has a hard, non-configurable timeout at 30s. I couldn't find in the docs whether *all* chalice deployments use it, but at least several of the standard deployments do. That makes the parameter in your [config](https://aws.github.io/chalice/topics/configfile.html?highlight=time):
`lambda_timeout`
irrelevant, I believe?

I'm not asking for a workaround at this point; I've abandoned this approach. I just want to change the docs to reflect this reality by mentioning the API gateway timeout, which is pretty non-obvious.

Thanks!

Start with the linked config documentation and its lambda_timeout section. Verify which standard deployments use the 30-second AWS API Gateway timeout, then update the documentation to explain that constraint and how it affects lambda_timeout; done means the limitation and its scope are clear.

Written by the indexing model from the issue text.