"I think if anything it has demonstrated that - using current space access infrastructure at least - the on orbit upgrading and maintainance of assets is hiddeously expensive and a complete false economy"

Presuming, of course, that we can continue to use the current infrastructure and modus operandi in to the indefinite future.

I don't believe that what we have now is all we'll ever have and we'll never do anything different. However, if it is in fact the case that we will never have anything better than the STS and ISS (which can actually be quite useful if liberated from NASA), then yes,

"(in most cases it would be cheaper to simply fly a replacement on an EELV)."

However, some of us are not terribly happy about continuing to throw away multi-hundred million dollar space assets.

As a banker I can tell you that a key reason our bank doesn't finance satellites is because if your satellite breaks there's no way to fix it. Broken collateral is worth $0, against an outstanding loan of perhaps a couple of hundred million. (I forget how much the Space Imaging syndicated facility was). Oh, and by the way, the debtor just lost their capability to generate revenue. That loan's in the toilet, and it's the depositors who suffer the loss in the form of lower returns on their deposits.

XM Satellite is a great example. Their Boeing sats are crapping out faster than expected. They had hoped to generate sufficient revenues to pay down the initial loans before going back to the financial markets to finance the #3 spare. That's been accelerated, and now XM has to go to the financial markets with a much worse balance sheet (lots of loans outstanding vis capital) and income statement (shorter amortization of the sats). That means they're paying more for the capital than they otherwise would have. Oh, and the insurance is going to be outrageous.

So is it necessarily cheaper? It depends a lot on how you're looking at it.

"That the Webb scope is going to a place where it can not be serviced is an excellent lesson learnt."

Actually, being at SEL-2 does not necessarily imply an inability to service. Sending a team out there to do it is pretty silly, but, the option exists to bring the Webb scope asset back to cislunar space.

All this really requires is an EML-1 station and the ability to maneuver trans-LEO. I believe the latter at least is supposed to be in the works.

I know for a fact that the return to EML-1 or LEO for servicing was considered as an option early on in the design process of the scope. However, the architecture didn't exist to do so. The likelihood of having the architecture in place in time was deemed negligible, so the option was dropped from consideration.

A nice current article on the whole concept, which was successfully used for the Genesis mission, can be found at:

Sciencenews.org/articles/20050416/bob9.asp

I still maintain that the ability to service orbital assets is a powerful tool. Without the ability to do so, Hubble would likely have been a billion dollar rock at the bottom of the ocean not too long after launch. With the human servicing capability it is beloved by billions around the world.