Move existing methods of inferring multi-montior support behind permission

[pes10k]

This issue is being filed as part of the requested PING privacy review #106

Currently its possible for scripts to infer whether the device has multiple monitors through the existing screen APIs. For example, screenX will be negative if the browser is on a secondary monitor, which is positioned left of the main monitor.

If there is going to be a "learn about multiple monitors" permission added to the platform though, this seems like a terrific opportunity to address some of this legacy-fingerprinting surface in .screen, and make the screen values only describe the current monitor, until the window-management permission is granted