As a participant, I can confirm the impression that the numbers were a problem. In particular, it was the size of the teams, rather than the number of teams, that was difficult. With such large teams, consensus-building was the biggest challenge, rather than resolving the options-cost problem itself. (I am sorry to say that our team principally learnt how not to do it, on both these fronts.)

So, one possibility could be to adapt the timing of the decision-making phases to the group size. There’s perhaps a special ajustment factor for the French to be taken into account, as well.