It just means that the minimum time between 2 allowed requests after your free retries are used grows similar to the Fibonacci sequence (the minimum delay for any request is equal to the sum of the minimum delays for the previous 2 requests).
In practice this interacts with minWait and maxWait so that the effective minimum time between requests after free retries are used is [minWait*1, minWait*1, minWait*2, minWait*3, minWait*5, minWait*8, minWait*13, minWait*21, ...] Increasing until you reach maxWait. In this way it's only fibobacci-like because the actual delays are not necessarily numbers on the Fibonacci sequence and because there is a maximum delay amount.
It just means that the minimum time between 2 allowed requests after your free retries are used grows similar to the Fibonacci sequence (the minimum delay for any request is equal to the sum of the minimum delays for the previous 2 requests).
In practice this interacts with minWait and maxWait so that the effective minimum time between requests after free retries are used is
[minWait*1, minWait*1, minWait*2, minWait*3, minWait*5, minWait*8, minWait*13, minWait*21, ...]
Increasing until you reach maxWait. In this way it's only fibobacci-like because the actual delays are not necessarily numbers on the Fibonacci sequence and because there is a maximum delay amount.