The detection framework is based on assumptions e.g. we sniff the agent and if it is this agent, we assume you are on a this device.
We have a database with common device user agents and Galaxy S3 is within those. It's quite unusual that you get 0 value.
As for desktop, that's actually expected, since it's literally impossible from the server to be able to know the dimensions of the browser. Not to mention, on desktop users are able to resize the browsers.
And on your other question: can the PageLayout / Detection framework be used in conjunction? Definitely yes. In fact you can use them to completely stop certain portions from rendering based on the various visible attributes.
Check out the Telerik Platform - the only platform that combines a rich set of UI tools with powerful cloud services to develop web, hybrid and native mobile apps.