The idea is not a new one - most of the first "digital cameras" were film cameras adapted with backs like this (although not quite as elegant a solution as this one). Interesting that they are still toying with the idea, although there are many pitfalls.
1) sensor. Full frame? It's going to be an extremely expensive, niche back. Protecting the sensor is potentially a problem - most digital backs are for medium format cameras, and are designed with the possibility of fitting a sensor in mind, and have the suitable precautions to prevent the sensor being fouled. No such case in a camera that was designed in the 1970s-1980s. Crop sensored? Would be cheaper, but you'd have problems with matching viewfinder to the sensor size, and viewfinder magnification is different on a whole range of different models which means each one would have to be bespoke, and either require masking or some sort of modification to the viewfinder.
2) Size - the design of the back plate on film cameras varied massively, about the only consistent design in the Nikon range was the FE(2)/FM(2) series, but there would have to be individual backs for individual cameras.
3) Cost vs. features - it's going to be expensive if it was ever brought to the market, but the small height of the back of most film cameras means that controls are going to be small, any LCD screen is going to be much smaller than the current sizes available on DSLRs, and places to put a memory card.
4) Power - film cameras were powered by a whole legion of different battery options - banned mercury batteries, 2x LR44 button cells, 3 button cells, 6V PX28, and everything inbetween. How would a back like this be powered?
5) Niche - it's a tiny market, and even amongst us that still shoot film cameras regularly, there's only going to be a small percentage of that cohort shooting Nikon, and an even smaller cohort in that group who'd want a back like this.
6) Film - believe it or not, and this is a point that is debated till the cows come home, but many people like shooting film and working with darkroom processes.
There's many more points that I'm sure go against this - I love the idea as a principle, but there are so many issues - many of which could be rectified with some elegant design work, but at huge cost - that I'm not sure I see the feasibility of it anytime soon. Now, let me go and load some Tri-X in the Rolleiflex...