The shuttle wasn't so bad for its initial run, but it has long since been time to scrap it and move on to the next project. With its 60's technology it's amazing that it performs as well as it has. Unfortunately while 98% reliability is good enough for research, it's nowhere near enough for production.

Time to break out the clean sheet.

I used to work as a contractor on the Shuttle and I heard a lot of stories about how it all originally came to be.

You see, the design isn't really what NASA wanted. They wanted a much smaller, cheaper Shuttle. They wanted a max payload of about 30,000 pounds, not the 65,000 pounds they got. The military insisted on the 65,000 number and got their way. This echoed through the rest of the design and spoiled a number of things (the Return To Launch Site failure mode is a joke, if the Shuttle is fully loaded, just to pick an example).

It was also an unfortunate time to do the computers that are on the Shuttles. These days, we can solve that whole fly-by-wire problem much more cheaply and for a lot less weight. The triply redundant computers that run the shuttle still used ferrite core memory when I was there in the mid 1980's - and not much of that.

The NASA bureaucracy is crazy, too. Lots of smart people are bogged down into the gooey center of an organization much larger than it needs to be.

I agree, the Shuttle is a mess. NASA is fine at exploration, but they are a crappy bus company. These guys have a better idea:

http://spacex.com/

and their first launch should be later this year!