There's a big difference between designing to fail and designing to be repairable. Even if it's possible it could have been cheaper to launch multiples but it's unlikely NASA could have gotten the funding to start making Hubble 2 in time for it to replace Hubble.
However, as to lifespan, from April 24, 1990 to Dec, 1993 the Hubble was significantly 'impaired'. If the next two averaged 10-12 years and the original continued to be useful for a few years that's well past break even. Especially considering the Hubble always suffered from a defective mirror even after the first servicing mission help that's not unreasonable.
PS: By designing to fail I don't mean they intentionally made it worse, just assuming servicing caused a wide range of problems. Sure, NASA did learn something from going though this exercise. But the highly political nature of the organization limits such benefits.
Edit: Yes, many telescopes had a very short lifespan though that was often by design. https://en.wikipedia.org/wiki/Rossi_X-ray_Timing_Explorer Mission duration 16 years, 6 days. https://en.wikipedia.org/wiki/International_Ultraviolet_Expl... 18 years. https://en.wikipedia.org/wiki/MOST_(satellite) 13+ (still up) https://en.wikipedia.org/wiki/Spitzer_Space_Telescope 13 years and counting. https://en.wikipedia.org/wiki/PAMELA_detector 10 years and counting. https://en.wikipedia.org/wiki/Solar_Anomalous_and_Magnetosph... 11 years 11 months. There are also quite a few up which may or may not last a long time.
There where 6 Hubble servicing missions. "this last Hubble servicing mission is expected to cost about $1.1 billion" http://www.space.com/6648-hubble-faq-space-telescope-repair-...
Further, 15 years is considered a relatively short lifetime for a satellite, so Hubble 2 and 3 could have easily had a longer total lifespan if they had kept them in a higher orbit because there was no need for servicing. Instead "the space telescope was designed to typically go only three years between overhauls."
AKA, they designed it to need 'fixing'.