Your point kind of proves the point of the GP: there is almost no middle ground. The differences between rates of success are orders of magnitude. Order N doesn't break even, N+1 almost makes you a millionaire and could effectively set you for life.
Some of the examples you give also had quite a history of a backing in some form, which is a totally different animal compared to a "I wanna make a game!" indie developer today with zero network and zero funding.