Regarding the best of 3 formats versus the single longer match format:
Kent Goulding and I did some research on this back in the 1990s when we were planning the World Cup tournaments. If you have a given amount of time to play, and the goal is to give the better player a larger chance to win, then a single long match is a much better skill determinant than a series of best n-out-of-m matches.
This is counter-intuitive to most players, so here's the explanation.
Suppose A is a 60-40 favorite over B in an 11-point match. (Pretty typical for a strong open-division player against a weak open-division player.)
If A and B instead decide to play a best 2-out-of-3 series of 11-point matches, A becomes a 64.8-to-35.2 favorite, a very small increase. To see this, note that A has three different winning sequences:
1) He wins two straight, probability (0.6)*(0.6) = 0.36, or
2) He wins two and loses one, which can happen in either of two sequences, ABA or BAA, each of which has probability (0.6)(0.4)(0.6) = 0.144. So total probability of this is 2*0.144 = 0.288.
Total probability that A wins is 0.36 + 0.288 = 0.648.
On the other hand, the average time required to play a best 2-out-of-3 series is about 2 to 2.5 times as long as a single match. In that time, they could play instead one 25-point match. A player who's a 60-40 favorite in an 11-point match is on the order of a 75% to 80% favorite in a 25-point match. (This estimate is based on a lot of match data collected in the 1980s and early 1990s when such long matches were fairly common in big tournaments.)
So, if you want skill to have a better chance of dominating, play a single long match.
However, most people think that best n-of-m series are more exciting and more fun, which argues the other way.
Hope this helps.