I don't know where you've find those formulas but they seem more complicated than the RHavar's code.

• X is uniformly distributed on [0; 1] because it comes from the seed
• The result is then divided by 100 to get the crash multiplier

So we can understand that we need to have (1-X) below 0.5 to get a crash multiplier above 2
99/0.5=99x2=198
=>CM=198/100=1.98

So we can conclude that less than 50% of (1-X) values give crash multipliers above 2 and thus less than 50% of X(=seed) values.