The latest Meteosat series, for example, is priced at some 2.2bn euros (£1.75bn)...
The natural desire for fiscal efficiency demanded that the four spacecraft built for the programme were all made at the same time (it's cheaper to mass produce), even though their launch would have to be staggered to provide continuity of operation over 20 years...
Of course, the consequence of this approach means you end up having a lot of very valuable, high-precision engineering sitting in store. Meteosat-10, the third platform to fly, has spent eight years on the ground waiting to make its ascent to orbit.
And the next bunch are due to cost €3.4bn, the first to reach orbit in 2018 and to cover through to 2040.
Even the article notes that because of the bulk manufacture these things are well out-of-date by the end of the process. I caught another article recently about how the US is investigating satellites that cost $500,000 to provide 'on-demand reconnaissance'. (Compare that to an article from 2011 where they say $100 million is cheap.)
That's the sort of money we should be talking about, half-a-million, throw it into space for a million, job done, it won't last as long but for that cost we could throw three, four or more up there and get better cover too. I wrote back in 2009 about companies that were trying to make getting into orbit cheaper.
I'm a little surprised satellite builders haven't followed car manufacturers and offers a stock platform with comms, propulsion, power, etc and just a standard interface to hook up the mission-specific payload (camera, instruments, laser, whatever). That could drive the cost down.
Now more than ever the UK needs to revive its launch credentials, we used to rock at it. There was even talk of it back in 2009, but we don't seem to have come very far. Somebody needs to get hold of it and drive it forward. Any of those Raspberry Pi people fancy a new venture?