Fix flaky logger spec

What does this MR do and why?

This spec was failing occasionally due to a slight time difference between the expected time and the actual time delay value for cursor_delay_s. I think a small amount of lag is acceptable so I've opted to use a range for the assertion (be_within(0.01)) as opposed to the fixed 0.0 expectation.

Most of the pipeline failures shown here #455157 (closed) were due to the actual value being 0.001 with the largest delay being 0.003 - so the chosen range of up to 0.01 should be acceptable.

Relates to #455157 (closed)

Changelog: fixed

EE: true

Related to #455157 (closed)