Well, take it from the perspective of a naive new cacher who does not consider the ramifications of placing a cache at location X. Deceangi has made it clear that as a GSP reviewer he's having to deal with people who have not even read the GSP guidelines so it's reasonable to deduce that they don't have any awareness of GAGB.
For example, my first cache placement was over-enthusiastic, placed in a spot where the imact of caching would have been highly detrimental to the habitat, although it passed review. It lasted less than a week before someone pointed out what was happening and I archived it hurriedly with remorse.
If my introduction to caching had lead me directly to GAGB because it's the de-facto first port of call for caching in UK then I could have got engaged faster, more effectively and avoided the blunder. I acted swiftly to rectify my mistake however not everyone can or does. I still made poor cache placement choices 9 months subsequently and the risk of doing so again has not gone completely, however, climbing the learning curve has been largely my own effort.
So, if GAGB were at the forefront of caching in the UK, collectively we could ensure that our knowledge is shared effectively, that caches placed are sympathetic and appropriate to the environment as well as the local circumstances pertinent to UK. Essentially, GAGB would provide the community and the information to help get new cachers to mastery of the craft in the shortest possible time. Mastery imples that they could review their own and other people's caches effectively.
I don't think the GSP reviewers would complain if the proportion of caches that come across their desks and are ready for publication first time round improves
I suspect their workload is a classic application of the 80-20 rule, that 20% of cache placements require 80% of the effort, so pushing that in the direction of 90-10 and ultimately 100-0 can only be a good thing. However, the benefit of achieving this through the GAGB is that it would apply across all the listing sites.