This debate regarding B/S and C/S makes no sense. If we have chance, why don't we find out a real case including a lot GUI fuctionalities that we might encounter commonly and implement it respectively by various GUI technologies, no matter what the technology is. Then, I suggest we publish result including source code, data base, time sheet and something else.
After all, we must put our idea into pratice.
Of course, the precondition is that we have enough time. If we have, why not? Banq, could you consider that please?