I find this thread very interesting and a good example of the reality that many people in technology face today.
Without going into to much detail I am involved with a handful of educational facilities that are trying to address similar challenges. To start off they are all higher ed so that does change things a bit and at this point each of them consider the potential of something like a Pi as a thin client to be a science experiment. Granted one they are very interested in but still very much isolated to the lab.
What I can say these organizations have learned through experience is the following.
VDI is an excellent solution in scale but it is less for financial than other factors like control, security, response times for new projects, mobility, and long term maintenance. Now each of these have costs associated with them but they are often discounted from a ROI when it gets to their finance people. The potential of a very low cost thin client is of key interest here because it can help tip the scale in the analysis of the costs that their finance teams really care about. In fact two of these organizations are looking at the exact same idea a 19 inch LED monitor and a usb powered Pi competitor that runs android with the VMware view client. Estimated total desktop cost of $230 and zero local desktop config if there is a problem they are replacing a cable, component, or SD card.
The next thing I would pass on is that it is all well and good to debate the benefits and use of the Microsoft stack but the reality in my experiance has been that anything outside of MS is pushing water up a hill. Of the 3 situations I know of where a education organization seriously tried to do this the student body blow back killed it. What I think is perhaps commonly overlooked is that it was not really killed by student preference the factors that changed the game were around external influences to the students lives. Stories about having issues exchanging files with graduate programs and potential employers. There was one situation that carried alot of weight where a partner organization that did internship placements made file exchange issues a big enough problem it got to the president of the college. Students and parents upset because recruiters told soon to be graduates that their inability to state they had MS Office experience kept them from being competitive in the job market. One I found particularly interesting was about how parents at a middle school revolted because of the "specialized computer requirements" the school had and how that complicated their lives. So what I have seen is that the reality of the software decision is much bigger than the organization itself and the comparison of technical capabilities. In my experience outside of developing countries where the decision is between have and have not most educational leaders would need to see enormous benefits to not using MS software to make fighting the battle worth while.
Now before any debate starts let me just say I am not endorsing these decisions nor am I aligned with one technology or the other. I do work for a technology provider but we are cloud focused and do not sell VDI or MS software licenses. My point here is just to help provide prospective and minimize any false starts. Technology projects are challenging enough so knowing what you might be up against is a big part of being successful.