This test tests timeseries retryable writes. It simulates retryable writes by issuing commands twice, both with the same lsid and txnNumber, and tests that the retriedStatementsCount and retriedCommands counts from the server are what we expect. 

However, in the rare care of one of the transactions actually being aborted and retried, the retriedStatementsCount/retriedCommandsCount will be one more than expected. This isn't an issue with the server code, since the statements are retried successfully, but the test doesn't account for this by checking for strict equality for the counts. This ticket would add a bit more leeway and make the test more resilient in the face of actually encountering a retried transaction.