I Never heard of a unique way of counting cycles for Williams so my comments are only general in nature
Cycle counts for most engine relate to fatigue life limitations of the parts under certification rules. They are hugely statistical in nature with typically margins as high as 100%.
The idea is to run a bunch of tests and in-service sampling to show that detectable cracks do not exist within the samples.
This transfers over to reviewing the operation in-service by counting only the landing cycles while assuming that 95% of your fleet never exceeds the rated life and that 100% of the fleet never exceeds an overage of a few percent.
If any part is found with detectable cracks the process needs review and adjustment of life for the whole fleet.