Could it be because of the LN testing?
Not likely. This is probably due to Segwit starting to be rolled out more and more.
LN is still a giant experiment at the moment, and I can't really recommend getting involved with it right now, unless you are into masochism.
We'll see how much of a difference LN starts to make once they hit 0.6 (0.5.2 right now)
We have less tx per day comparing to the previous weeks. *Maybe* that's a good explanation after all.
https://bitinfocharts.com/comparison/transactions-btc-ltc-eth.htmlIt's number of outputs what is most relevant here. During the past few months many users, exchanges, etc started to consolidate addresses and batch transactions to reduce the impact of the high fees. That consolidation process plus the numerous forks we experienced that required to move coins before claiming made for an abnormal higher number of tx's which added to the already congested situation.
Now, after that consolidation seems to have been stabilized, more tx batching adopted, segwit increasing adoption, , no more forks to claim, etc... it is completely normal we reach the benefits of an uncongested network.
And of course, the increasing adoption of Segwit allows for more "space" available even for non-segwit tx', as expected.
We don't even need to take into account if spam was or wasn't a significant factor to see the "optimizations" in usage could perfectly explain the improvement.