I am very involved, and I've seen plenty of good work, from Java 8, through the upcoming Valhalla and Panama, and to the amazing work on Graal, which is probably the biggest breakthrough in compiler technology in the past decade. Doug Lea is still working as tirelessly as ever on java.util.concurrent, and while letting the evangelists go was bad PR, it really changed little wrt to technology investment. There have been some missteps and setbacks, but while making some noise, they have been few. Don't forget that Java is huge; probably the largest standardized language ecosystem ever, and not all of its many corners can always be covered well. And when an ecosystem is nearly ten million users strong, even one percent of disappointed users is larger than most other languages' entire ecosystem.

This may well have been a mistake, but it is being addressed now, though probably not aggressively enough. Also, depending on how Google integrates OpenJDK into Android, we may soon have Java (real Java, that is) on the world's leading mobile platform.