Feature request - Add Retry-After header when the API returns 429 status code

Our services occasionally receive 429 Too Many Requests errors when we reach concurrency limits. A specific example we encounter is:
Project exceeded maximum number='8' of concurrent instance_read requests. Please try again later.

Currently, to handle these situations, we rely on a client-side exponential backoff strategy. While functional, this method requires us to guess the appropriate delay, which can be inefficient. It can lead to our services either waiting too long or, more critically, retrying too quickly, thereby placing unnecessary load back onto the API when it's already at capacity.

To address this, we kindly request the inclusion of the standard Retry-After HTTP header in all 429 responses.

We are confident that this small addition would be a significant improvement. We would appreciate it if you could consider this feature for your development roadmap.

Thank you for your time!