If you look on evidencebasedbirth.com (excellent site) it says that up until recently it was thought that risk based management (UK method) was as good as the routine screening method. Recently, they have found that routine screening has a slight increase in success rate over risk based management.
The main issue with the routine screening is that GBS is not present all the time. Routine screening can still miss infections and can alert to infections which have passed by the time the woman is in labour. If there was a test which was fast enough to be done during labour, that would be the most accurate way of testing. At present, we don't (but we can use it to assess the accuracy of tests done before labour.)
However according to the EBB article, routine screening does look to be more accurate than screening only women at high risk.
(Is the screening not just a urine test?)
I am anti unnecessary interventions (not ANY interventions, clearly, that would be ludicrous. I don't think anybody else on this thread is arguing this at all.) but if they have a well proven benefit, which this does seem to according to fairly recent US trials, then I would be for it.
To reduce the impact of the medicalisation aspect (which is a problem, and as ElectraCute put it excellently earlier in the thread: "it does have implications beyond some arbitrary 'fear of medicalisation'... they have nothing to do with airy-fairy fantasies about natural birth.") a cannula can be used which can be detached from the IV when the antibiotics are not being directly given, which allows for freedom of movement, use of water pools, etc.