When grouping by a numeric value (think "http_response_time_ms" or something similar), it would be handy to have a way to bucketize the results, otherwise we wind up with a result for every unique response "http_response_time_ms" value, for example.
I'm not sure this is the best API suggestion, but maybe something like "group_by_each=100" which would create a bucket for every 100 ms.