Do you have a good source to point to where I can go read the technical aspects of what manipulation were performed. It is not unusual to have to make choices on data - particularly when it conflicts. If one data set makes more sense with all of the other data when put together you may give it more weight and choose to include it. It is not unusual to include the rejected data in supplemental material along with an explanation though. But some profs don't like that approach ... I do.
On the other hand, taking portions of a single data set and rejecting others when all the data was collected in the same way and with the same presumed accuracy is an issue. That is what I would more classically call cherry picking.
Examples along with technical descriptions of how the data was screened, processed, rejected, etc would be helpful because I didn't see a lot of instances that caused me a whole lot of concern on my first go at it.
Most of Michigan is on eastern. I made such a habit of short nights for my first 30 years, I haven't found a way to break it yet in my second 30 years.