I do not think that it would have any impact on amount of received traffic. Or do you mean the amount of IOPS?
It may has an impact on duration of the used-space-filewalker scan though, but I do not think it would be significant.
It would be nice if you could compare the time for the used-space-filewalker with the badger cache enabled on the HDD and on the SSD for the same node but with a several hours difference between attempts (to invalidate the cache) and publish them there: