### Feature description - Max throughput would prioritise throughput over latency: enable batching, use refetch delay (see: #500), etc - Min latency would prioritise latency from filing a job to executing it: minimal batching, no refetch delay, etc - High scale might be a balance of these suitable for people who are dealing with very high numbers of workers/jobs. ### Motivating example People don't like to read documentation or have too many options; pick some sensible ones for them. ### Breaking changes None.