[Metrics] Consume measurements of remaining async instruments when one fails

[overmeulen]

Describe the solution you'd like
Today when the callback of one async instrument fails the whole iteration fails and the remaining measurements are skipped.
I'd like for the exception to be caught and logged and for the remaining async instruments to be processed normally

Describe alternatives you've considered
Today I'm handling this on my side by decorating my callback to catch and log any exceptions

Additional context
https://github.com/open-telemetry/opentelemetry-python/blob/main/opentelemetry-sdk/src/opentelemetry/sdk/_metrics/measurement_consumer.py#L74

[aabmass]

@overmeulen thank you for trying out metrics, this this is indeed a bug.