tablets and mobile devices are designed for portability and, not too different to most net/ultra books being shipped today, this means minimising space by soldering essential components onto themain board (CPU/GPU,Ram, flash Storage etc.)

This makes it virtually impossible to buy individual components without economy of scale via a proven manufacturing capability.
Large CPU manufacturers that use Arm technology, for example, place very stringent demands on manufacturers to reveal technical and design information about the sort of products they will be putting their hardware into (to ensure quality and performance and brand consistency).

This is part of the reason why so few "quality" open source tablet type devices have been made/sold without being to a "preordained/determined" spec as set by the manufacturer.

If you've followed the scene, you will find that most people who want to design/develop an open source device, start out with a high spec concept and then end up with whatever low-mid spec system they were able to actually get their software working on.

I genuinely believe that building on an existing base is the best way to go. the Iconia w500 can run Mer and thus technically there is no issue with running Hildon UI on top of it (or any UI for that matter. With Vivaldi so close to market, why go through all the heartache (and delays) that they have, just to create a parallel piece of hardware, when others have done it for you?