Hi all,
I'd like to add another voice to this request, and some comment.
Firstly, I work for a large retailer - we would love to manage our signage in Revit, but cannot due to the lack of things such as face-based parametric materials (ie, auto-scalable materials that react to the size of the object) as the deeply-buried image parameters are simply not accessible. We have about 2,500 different pieces of signage to manage, and the time it would take to hand-code these in Revit is excessive.
Secondly, while in theory Revit / 3DS Max share the same texture libraries, this is not strictly the case. What-you-see in Revit is NOT what-you-get in 3DS Max. Whatever the shaders the Revit uses, these perform horribly in 3DS Max, which means we rely on 3rd party packages such as vRay or 3rd party conversion scripts to make the Revit textures usable. In particular Max's performance with Revit procedural textures is abysmal, as Max appears to convert the procedural textures on-the-fly to temp raster textures on a face-by-face basis. This results in thousands of temp textures that again vastly slow the render process.
There is no clear, straight out of the box workflow from Revit to 3DS Max that gives the same render image when doing a basic Revit -> FBX -> Max -> Render. My interpretation of 'sharing the same material engine' (and indeed rendering engine) is that such a simple export-import-start render functionality should be possible. In practice it takes a *lot* of fiddling to get Max to produce a similar output to Revit.
That said, I would prefer to do all rendering in Revit, at least for stills; the cloud-render capability is a massive time-saver in this regard, and Autodesk is to be commended for getting this aspect working as well as it does. However we are strictly limited by the inability to place the full singage for a store in Revit, which means our renders always look 'wrong' - the signage adds a huge amount of colour to a store. Yes, we could use decals; but it means placing hundereds per store by hand, that are not able to be scheduled, nor can they have the kind of BIM metadata we need to take from the RVT model (order codes, manufacturer, cost, etc.).
For us it is this BIM aspect that is the killer: the signage is the one component of the design that we cannot (yet) easily bring into the world of BIM - at least not if we want the signage to be WYSIWYG.