Just adding to my last post, lf you read the patent I have listed, it doesn't take much to realise that blaming the Mylar surface is an easy way out to establish a reason for failure.
The patent, when carefully studied, clearly indicates the immense importance and methodology applied to clamping the film and the processes that follow.
That ''perimeter'' which comes in to contact with the stretched Mylar is where many companies that have spent large sums of money in R&D on this idea, have NOT taken in to consideration.
This, in conjuction with the very thorough adjustment points to flatten the film before a vacuum is applied is what makes this design work.
Stretching Mylar over drums, bicycle wheel rims and other stupidity will indeed produce only a ''shaving mirror'', if that.
I have been told in person by someone involved in this project all those years back, that the above mentioned ''tricks'' is what sent many researchers and developers in to bankruptcy.
I was also told that holding a vacuum is not diifficult at all with ''just a thin smear of silicone grease'' and that ''a simple regulator had easily been deployed at very little expense to hold the vacuum constant''.
We can sit here and **** can this as much as we like, but read the following link, about half way down in particular, where a paragraph starts with ''In 1985''
http://spie.org/x26677.xml?ArticleID=x26677