There is not. Identify is a fairly expensive operation and might take several seconds if the layer requires a request to the server.
You could do it one-by-one from top down on just the layers that are rendered client-side, but you will need to have code that stops the process once the mouse moves again, so you don't keep identifying on locations you no longer care about. However it'll still be a somewhat expensive operation, as it requires an entire render-pass each time around the area of the pixel that you're identifying on.
We are looking at if it's possible to add a more efficient "mouse enter" and "mouse leave" event for a few layers in a future release.
