Logging just the function signature is not very useful when analyzing mint failures - an additional description of test / naming test descriptively will help (e.g. "PutObject test with custom metadata")
Currently many SDK's log the errors as PASS/FAIL with no meaning error message associated with it.
SDK's stop on the very first failure and do not proceed further. A better approach would be to run all tests and raise issues for failures against the respective SDK/Mint so that they can be tracked and resolved for faster iterations. Playing catch up will be really slow when Mint and SDK updates occur at tandem.