This reminds me of the engineering adage, "you can always get one to work", meaning you can always tweak the engineering prototype to work in a controlled laboratory setting. Trying to get a lot to work identically, i.e. get it into production and crank out thousands of identical units, is the real trick. Likewise, hosting a movie in select theatres and extrapolating success based on a few data points is pure BS.
It doesn't need a large audience. Heck, I don't know why people are so surprised that people are seeing it. They go to Merchant Ivory films, don't they?