I started out arguing against COL, as an incredibly expensive and inappropriate initiative. The dumbest feature is that public money pays for things in ways that encourage the producers to distribute them narrowly.
Something similar happens with many LEAs that use their share of Standards Fund grants to buy things, then hide them behind passwords - which keep out even their own people. I think that's immoral. In the East Riding, I have won that argument, and now we put everything out to the world for free (enlightened self-interest).
Far better is the National Learning Network route - the NLN stumps up the DfES cash. The publishers tender to produce materials to a very high spec (including excellent accessibility guidelines). Then they commission the writers. The stuff they produce then goes into the public domain. It's not that easy to find, as NLN mainly serves FE colleges, but once you find it, you get it for free.
www.nln.ac.uk/resources.aspYou may need to claim that you belong to a college, but just pick one near you, or one you know.
But I have come to realize that COL is a device to sustain publishers and subsidise them in a market, where purveyors of freebies (like John and me) have made their online activity less attractive.
COL has become horribly entangled with its own rules (like the 80:20 one that is becoming impossibly arcane). It would probably make more sense to let schools have discretion in what they spend, but use other means to steer them away from spending it all on more Bill Gatesware. Or maybe go the other way, like the Catalan ministry with CLIC, and pay for the development of tools and content, then give it to the schools (and the world).
There are some publishers (definitely cottage industries) that have not been able to jump through the COL hoops (Question Tools, for example). And for a long time COL would not recognize teachers or anyone without business accounts. (I am proud to say that I kept naming and shaming them for this, until they gave way last summer.) Cottage industries rarely have the time to sort out the COL rules, the metatagging and so on. But the big publishers - RM, Granada, Learn.Co ProQuest - they can put millions into this, as a speculation, then let the money roll in.
The commercial publishers want to sell subscriptions, which lets them rest on their laurels. With our LEA's (meagre) money and our regional grid's (a bit more) we pay local programmers to develop stuff, that we then own. If they do well, we will keep paying them to do new stuff - but we own it, not a commercial provider. And we provide much-needed employment in Yorkshire, rather than London or Oxfordshire...
I've recently coordinated a pan-European ICT project, which is part of a wider validation measure (ValNet) promoted by the European Schoolnet. Among various findings is one that supports Graham's view: teachers in England, Norway, Catalonia, were not short of suitable software (and had no interest in some rather clunky freebies that they were offered in the project). What they want is time to use the things they have already, to develop learning objects, and simply to become confident in using the software for teaching and learning. Oh, and they want a lot more hardware and infrastructure - more networking, more bandwidth...