Tulip, I look at things differently than most but I agree with Tim C. in that FF is really nothing more than teaching the dog to be obedient about fetching and ultimately going when sent. Where I think some of the opinions about the "larger effects" come into play is more or less rooted in what one might call "traditional" types of "programs" where FF is the first place the dog is introduced to the concept of "turning off pressure". In those programs, they may have learned to avoid a pop on a choke chain but they probably don't understand the concept of continuous pressure being released when a command is performed until somewhere in the "force fetch" process.
I train dogs of all shapes and sizes and most often get called for large dogs that have had some basic obedience training (rewards only) but are out of control at the adolescent stage, around 6-12 months. The vast minority of these are retrievers and other than my own dogs I don't think I have ever "force fetched" a one of them bu I frequently employ things like prong and e-collars to help owners bring these dogs under control. What I consistently hear back from people is that after a short time in training their dog becomes a new dog, meaning the unruly youngster, while they may still have their moments, is much more quick to listen up and do what they are asked.
I've often pondered why that is, in the vein of trying to improve my protocols, and I think it has a few components:
1. More consistent reinforcement
2. More frequent reinforcement
3. More kinds of reinforcement (rewards + turn off pressure)
4. Higher expectations leading to higher standards from owners that ultimately lead to less practice of bad behavior
I think in long term (advanced) retriever training we often believe we are looking at dis-obedience, when in fact we may be looking at a lack of confidence or the dog being confused (as Tim mentioned). There are cases, most certainly where a dog has not decided to follow it's handler because of weak reinforcement in previous training steps or a discontinuation of reinforcement as the dog became reliable. In a lot of cases people stop reinforcing good behavior as dogs become reliable and then later these behaviors can weaken. I see that all of the time in the pet dog practice too. In fact, it is the real reason I get called for all these adolescents. People get tired of smelling like liver treats all of the time.
I can remember a great case in point for this and I wish I could remember the dogs name but he was a bull of a BLM I encountered in my work as a retriever trainer. I had come to a new team and when I did the team members instructed me that we always ran this dog last, especially on blinds. Seemed he had a habit of bringing back one bumper on the first blind, two on the second and three on the third, which if you've ever worked 20 dogs in a row can be very annoying. Everyone thought he was kind of cute that he seemed to be able to count but in watching him.. he was dog aggressive, would pee on every bush when returning from a blind, aired whenever he felt like and generally carried himself with an air of "oh yea, what you gonna do about it". He pretty much did 80% of his stuff on his own terms, at least in his mind. No discredit to the team for managing this behavior instead of addressing it. Dogs were at a premium right then and going into intensive therapy for this ran the risk of the dog getting screwed up for a while, which we couldn't afford. I was able at the time to get the team leader to consent to letting me work with this guy between training exercises so I took him into obedience and basic pile work, where I upped his standard and made sure he didn't ever shop the pile but rather returned straight to me with the first bumper he came across. You could say I went back to "force fetch" or "force to pile" with him but I really didn't, I just used those exercises to show him that if he screwed around there were bad consequences and if he did it my way, it was a lot of fun to work with me. It was never that he didn't like to retrieve, just that he didn't like to be controlled with inconsistent handlers (there were a lot of people handling him at this time so total consistently was impossible).
It turned out in just a few sessions that this guy stopped shopping piles and peeing on the return, became more obedience in general and even started taking his casts better on blinds. He transformed pretty quickly once he had a clear set of expectations in his mind and consistent enough reinforcement to habituate those good behaviors. He was a great lesson for me in the "larger effects" of force fetch which, as I mentioned, I believe have nothing to do with "force fetch" but rather to do with consistent reinforcement of the behaviors we want and punishment of those we don't.
That was a dog someone would say "put him back through FF and the yard", and truly, that wouldn't have hurt this guy but it wasn't really necessary once we got an understanding.
So long way around but I think i advanced retriever training literally any behavior can and will get weakened by lack of reinforcement, requiring the trainer to go back to some point in the training process and rebuild/reinforce that behavior more consistently. Where exactly they "back up to" would depend on the behavior that is suffering but I think a lot of things get generalized back to "weak FF" when in fact they are just behaviors that haven't gotten enough reinforcement lately to keep them reliable.
So, given FF is really focused on four or five behaviors (go, pick up, hold, here, release) I think those are the ones that necessitate going all the way back to the beginning of the process however, as with the example above, any of the steps might be used to resolve a particular behavioral issue later in life.