Tall pine, you are right, with one caveat, it's notthe refit recurring cost (which is actually quite low compared to the other costs) that is the problem with the shuttle, it's the yearly infrastructure and personnel cost--when that is amortized over only 3-5 flights a year, instead of 10-20, it renders the shuttle too expensive...which dries up the market (kind of like the non-recurring costs of RDTE that result in procurement of aircraft being reduced--see my previous posts on that topic). The problem was, the market wasn't there at that beginning (other than non-private missions), so the costs ended up being higher, which further reduced commercial interest. Now, the commercial market is growing (and has been, just not as fast as futurists believed over the past 50 years), and will reach a point where it coukdsupport a reusuable system...I would say 2020-2030 timeframe (when the market is sufficient to allow for economic operation of a reusable that would compete with current low cost expendables). The shuttle was just too early, and too reliant on government missions to make it's flight count. With the post cold war defense and intel drawdown, combined with the challenger and Columbia accidents causing delays, the market shifted to smaller, lower cost systems. That shift made the per flight price too high.
There are other examples of too-early concepts in space systems (iridium and teledesic for example, which would have been far more profitable and/or gotten off the ground had they started 10 years later).
That being said, the companies interested in long term profitability of launches (spaceX for example) do have long term plans for reusable systems or derivatives as they do recognize the above.
It's similar in a way to installation of fiber backbone connections..,a huge amount of fiber was installed in the late 90's and early 2000's, anticipating a huge and rapid rise in bandwidth needs...due to the .com crunch and subsequent 2007/8 crisis, the need was slowed, prices collapsed, and many investments lost severly, and there was a bandwidth glut...that low cost bandwidth actually did lead to a "field of dreams" result (if you build it, the demand will appear), and it did (cloud apps, mobile backhaul, streaming media), and those apps now are causing a bandwidth crunch (and higher prices, and more profits for those who bought the bankrupt lines for pennies on the dollar)...now the high demand and associated prices (which are actually the market correct prices) are motivating further installations.
Space is similar in a way, except the demand rise is much slower, and the leave-behind infrastructure that can be used limited, so the collapse/reuse/demand drive doesn't occur, meaning it's the long term trend that is needed to create the market.
I believe as the mobile market continues to expand, and spectrum becomes more scarce, and mobility needs rise (more remote areas) the LEO comms market will continue to expand, and that will encourage more space demand, but it is years out. Which is why launch companies are working toward profitability at current demand, not predicted demand, and leveraging that success toward being able to expand into high demand areas when it materializes with reusables (spaceX). Ironically, that company, for being founded by an internet guy, has been the most capable of avoiding an Internet or computer-like business strategy (assume exponential demand growth).
In any case, I do even more firmly believe that we will get there (increased space access, etc) eventually, in 10-30 years (depending entirely on how our politics either hastens a collapse, or corrects the current path and encourages true economic growth). 1990-2020 I believe will be looked back at as the space dark ages, where external influences (politics, etc) set back technological development and we had to rebuild from a point more similar to a time 20 years before the dark ages, than simply continuing forward. (as an example, ares uses a derivativeof the J-2 engine, which was the Saturn-v 2nd and 3rd stage engine, instead of more modern engines...due to a risk averse design). True progress requires risk, governments, by definition are risk averse, while correspondingly entreprenuership also rewards and requires calculated risk...that's why private demand and private supply will be the future.