It looks like the James Webb cost $8.7B to launch. The Hubble $4.7B at launch and $10B through till 2010. So it seems cost is the limiting factor. Replicas might be cheaper, but they wouldn't be kickstarter cheap.
I'm under the impression that R&D is a significant component of the cost in these sorts of "one off" science projects/devices and consequently, manufacturing, launching and operating a second one might not cost anywhere near as much as the first one.