"In fact it's hard to find a single shuttle mission that has gone off according to the ideal of the project for a fast-turnaround, reusable, reliable and versatile vehicle."
In fact studies *do* exist of (AIAA-1995-3527) Shuttle launches. It's not *quite* as bad as you think (but it's not outstanding). During 1981-2000 of 204 launch attempts, 50 went off on schedule , 111 were "scrubbed", 17 were < 1 week late, 13 < 1 month late and 13 > 1 month late.
At a price of roughly $1m a day. Each scrub was costed at $616k.
Lowlights of this was the "Summer of Hydrogen" (5/30/90-10/6/90) roughly a 100 days due to a Hydrogen leak in the airtight engine compartment (so instead of a little flame you'd get a confined explosion, like the difference between a firecracker on your palm and wrapping your hand around it). It seems LN2 (BP roughly 4x that of LH2 and 10x LH2's viscosity) which is what the engine makers used to leak test it, makes a poor substitute for LH2.
That said it *could* have been a lot better either in design or during upgrades (especially given the money NASA spent on designing but not *installing* various upgrades). Even so it did manage on time take off c25% of the time and gave the US a new very high efficiency LH2 rocket engine of *known* performance. That's a *very* useful bit of hardware to have *provided* you want to build another LH2 fueled RLV, as they're just too dam expensive to expend.