It's not mutual determination, it's that the definition of "capital intensive" can switch when the interest rate switches. That is, if the interest rate is above 10% per year, it may be that production process "A" is more capital-intensive than production process "B"; between 4% and 10%, production process "B" is more capital-intensive than production process "A"; and below 4% production process "A" is more capital-intensive again.
In such situations, the claim that lower interest rates trigger shifts to more capital-intensive production processes is incoherent...
Of course, there are similar problems in garden-variety consumer demand theory: Giffen goods, where (over some range) demand increases as the price rises.
The big question has always been: "Are the Cambridge counterexamples unlikely freaks of theory, or do they say something about the situations we are likely to find in the real world?"
On this question I am agnostic leaning toward the MIT view: no one's shown up with an empirical example of "reswitching", or even a real investment project with a cash-flow profile that would make "reswitching" likely.
On the other hand, there are big and deep problems with the Solow production function view that are completely unconnected with the Cambrdige Controversy...
Brad DeLong