One for all and all for one 6

With three players, each participant's role is more valuable than with five.

Undoubtedly, that's why the team compositions turn out to be more unbalanced, and the strongest captains have a clear advantage. Last time, there were 16 teams of 5 people. If there were 3 people per team, there would be 26 teams! Is it necessary to explain that the difference in skill within one group would be significantly more pronounced in this case than with 16 teams? Not only will the variance between the first and last captain be large, but the same picture will be in one of the groups, meaning that the weaker captains will initially have fewer chances.

As for the example with your team, I don't think it's representative. My team was one win short of the semi-finals, but I believe that everyone played as responsibly as possible, and the number of people on the team somehow didn't hinder us.

And Cobra is right: in a team tournament, you need to choose players not only based on skill but also on reliability. No matter what the replacement system is (and I'll think about how to organize it best before the start, perhaps we'll take a variant from Game of Thrones), any replacement will still weaken the team (it's important that the replacement doesn't strengthen the team, otherwise what's the point of the draft).