I want to write a book called The Mythical Monkey-Month. It will argue against the apparently common illusion that, given a language with such enterprisey features as static type checking and compile-time provability, you can hire barely-trained monkeys and solve actual business problems well.
I do believe in correlation between the capabilities of a language its environment and higher levels of productivity, but bitter experience has taught me that there are no technical solutions sufficient to protect against incompetent or at least malicious developers.
(See the Perl Monks thread Perl in the Enterprise to play logical fallacy anonymonk bingo within this subject area.)
