Hand in example is GF(3NT) statisticaly. There is low chances for void
♣, so you make 7 tricks on finesse
♣. Even with light opening your p have statisticaly 10hcp, normaly AKQJ scattered in suits. This mean 1 sure trick for A and 1 more when you take opening lead with your stopper. But question is how to bid it, of course preferable by p, because he hold stoppers.
It is impossible to do this playing 2/1 GF without some lie, because 2/1 system sacrificing inv part of 2/1 bids, is very uneffective in most common case of inv responsive hands and lead often to wrong sided NT contracts after 1NT RF. Even clear 2/1 GF bid, which rarely comes, actually does not bring any advantages. Because of above weaknesses you need also ton of gadgets to avoid it and lead to unnecessary complication of system to about 300 pages.
I will illustrate my thoughts by given in post example:
1
♠-2
♣(only possible nat GF bid in 2/1),
2
♠(most common rebid) - 3
♣(without lie only possible bid)
?
Without gadgets continuations here are very unclear. Depending of reponder's intentions bids of opener must be different. If 3
♣ establish
♣ suit and show slam try cue bids and serious/frivolous 3NT apply. If 3
♣ was for finding best game, then bids are stoppers and deny stopper in unbid suit, which is case in example. Because these 2 possibilities contradict each other, you again need some gadget...
What is the solution? IMO - give up playing nat GF responses, switch to transfers and one GF relay
Misho