There is no inventive for researchers to break new features before they are released. In fact it is often more trouble than it’s worth to point them out.
On a related note, some folks have started giving pre-launch advice on Chrome's "intent to implement" threads: https://groups.google.com/a/chromium.org/forum/#!topic/security-dev/2H3ZiI_INm0 … It doesn't solve the incentive problem, but might be worth following and/or chiming in out of the goodness of your heart :)