I suspect this is already on the roadmap for the Find-It team, but I haven't seen a crbug specifically targeting this. Please dupe if this feature request already exists.

In this crbug:
https://bugs.chromium.org/p/chromium/issues/detail?id=908517#c3

Find-It detected a flaky test, but no one followed up [probably because the flake is reasonably infrequent]. I independently found the test. I ran the test locally and discovered that it deterministically fails when run by itself. Presumably it passes when run in a batch due to state being carried over from previous tests. 

These failures are very likely to cause false rejects, and significantly increase run-time.

Example series of events [matching the events from the above crbug]. Numbers are fictional.

(1) Developer introduces a test. It passes 99% of the time when run as part of a batch. It fails 100% of the time when run by itself.
(2) The test is run once, in a batch, and passes the CQ. It is landed.
(3) Subsequently, ~1% of CLs fail the test when run in a batch. Both 'with patch' retries and 'retry with patch' will deterministically fail, since they run the test by itself.
(4) We then retry the build at the CQ layer. Since we're retrying all tests and the bad test is run in a batch, it most likely passes.

In the example above, I manually ran the test locally, confirmed that it deterministically fails when run standalone, and then disabled the test. Ideally we could automate this process.