Jobs fail with Too Many Requests when attempting to submit AWS Batch jobs

[asjohnston-asf]

When submitting a large number of jobs at once, some job step function executions fail when submitting the AWS Batch job with:

Too Many Requests (Service: AWSBatch; Status Code: 429; Error Code: TooManyRequestsException; Request ID: 8c6b2219-1c07-4c9d-bb18-99facd090b01; Proxy: null)

The start-execution-manager lambda invokes start-execution-worker with up to three batches of 300 jobs. The worker lambda became more efficient earlier this year; runtime for 300 jobs is now around 20 seconds. That occasionally pushes us over the 50 SubmitJobs calls per second service limit.

For more context, see #1272