During our Passive House training, we were told recent results are beginning to reveal that the best location for a window in the wall assembly is in the middle. The isotherms (lines of constant temperature in section views) stay as straight as possible in this situation - their curvatures to meet the window glass from wall edges is minimized. Reduced curvature in isotherms generally means reduced heat loss. One might even say that the distance between isotherms is proportional to heat loss (flux), but I think it is more complex than that. In the case of the vacuum insulated panel, the isotherms would be very close together simply due to the form factor, (they have an R-rating of about R-50/inch in IP units), but the heat flux would still be relatively low. These details need elaboration, but I don't know much about it yet....
So are Deep window wells inside Passive Houses a good thing or not?
I have been fiddling with my PHPP and reading the 'research' and find I don't find a convincing argument that they are a bad idea, which I think some people are apt to say. Very crude experimentation with the PHPP will show that window reveals are of major importance in determining the solar gains available to the building. Thus, placing the window near the exterior can have major benefits for overall thermal perfomance of the building, despite less favourable psi-install values. Note that the window configuration is of significant concern here. Where you have a very large window surface, with lots of glass compared to the perimeter, the importance of the reveal depth is reduced, and the psi-install value may be highlighted as a potential focus for further heat loss reductions. This may also be the case when the wall is not very thick. But when you have a very thick wall (say 24" or 0.6m), and the windows are not too big (say 40"x72"h at most), then the depth of reveal plays a major role in the building's performance - I would say a larger role than the psi-install value. Therefore, we can still sometimes have deep window nooks in Passive Houses in cold climates.
With respect to isotherms, you will find that the corners where the walls and sill and heads meet the opening can be chamferred. The isotherms at these locations stay well away from the vertices anyways, so chamferring the opening where the door is installed, for example, should be fine in terms of energy efficiency.