Our problem with vists is response. We started off doing unnotified visits, but only found the claimants in 40% of the time. We then changed to notified visits, where we wrote and informed then when we would be visiting and the sucess rate (where the claimant was in) actually fell to 32%. The claimants only seem to get in touch when we suspend or even cancel their claims.
However when we actually get to visit we have found the visits to be very effective, as there is usually a change in circs (or six) in the vast majority of cases.
But then my borough has a very large percentage of working claimants (second highest in the country), who are nearly all receving tax credit. Its is usually a certanty that the tax credit will have changed, even if nothing else has.
But then we could probably pick these changes up just as easily with a postal claim form.