Well, got AC with heuristical formula based on generated tests for V=1,2,3. Got nothing to prove it anyway.
By the way, there is some lack of logic in the problem statement. First is that you actually have to output achievable score, not just minimal value greater than previous standing. This is arguable.
Also, as long as the team knows its score, it also knows outcome of games it played to achieve this score. This puts some limits on game outcomes to consider. E.g. if we wonder if our score of 3 might be winnable in some case for other teams, we already know by that time how we achieved this score. Either by 3 wins (so no team will gain extra point in case of 4 teams) or by 3 draws (all 3 other teams will gain one extra point). But we check all cases even for our own games, like we know our score after playing own games but we know nothing about goals we scored and even which games we have won.
This correlates nicely with:
we are interested to find the minimum point that a team can score and be lucky enough to qualify.
But this phrase:
This 5-point difference is the minimum bar that a team should reach if they are to have any hope of moving on.
sounds like they will definitely become qualified based on games between other teams. The wording "they are to have any hope" implies that they use all the information they have. And games played to get the score is part of that information (not provided in the input).