Almost all uninteresting...
Was skimming that list and, like most Java releases, it's the usual round of internal changes and syntactic sugar. Nice for Oracle, not so useful for developers.
But then I saw the Vector proposal! Yes! Now this one has potential - we do a lot of image processing in Java, and this could definitely lead to speed improvements. I investigated the HotSpot auto-vectorization a few years back, things need to be done just-so to take advantage of it. Making it explicit, with a non-vectorised fallback if the architecture doesn't support it is a great idea.
As for people "adapting to the accelerated release cadence": bollocks. We don't even test on the non-LTR releases because none of our customers use them. I'm not sure which "Java-oriented enterprises" have decided to adopt a JVM that's only supported for 6 months - the ones we deal with would take that long just to make the decision to upgrade.