Designing hospital wards

Article in NYTimes about hospital design:
Health Outcomes Driving New Hospital Design - NYTimes.com

I worked on a study where we had to provide input on the benefits and trade-offs of having more single beds in a unit. This article seems to suggest that all single bedded units are good. The argument is that patients have increased privacy and less hospital acquired infections. Also, private rooms do reduce need for transfers, and can increase your room utilization.
When we looked into it, there was no consensus on what percent of rooms should be private, but estimates ranged from 50 to 100%. Some arguments against 100%:
-Obviously, the more you move to 100%, the larger your estate needs to be - leading to higher costs
-Private rooms have been shown to prevent hospital acquired infections a bit more than shared rooms, but one of the biggest driver of hospital acquired infections is health care workers following standard precautions, not the sharing of rooms
-There is little data to suggest that length of stay reduces for patients in private room
-Some patients in private rooms tend to feel more socially isolated
-It is harder for nurses to survey their patients and prevent injuries

I'll be the first to admit that these are not the strongest arguments against 100% private beds - I think the argument is strongest in an ICU where there is a tremendous amount of equipment and patients are critically ill. But in a general ward, I think the argument is less strong. And there are cheaper levers than redesigning a hospital from the ground up to reduce hospital acquired infections and make patients feel happier during their stay. Not arguing against 100% single bedded units, but as in any discussion like this, it's a cost-benefit analysis.