The high crossover rate is not unblinded data. Of course they know how many patients are IN the trial and will be control arm or not. They are not required to be so blind that they have no idea what they are doing in any possible way. That would be absurd.
Without breaking the blind the crossover rate could not be known. All 331 patients were given the option to cross.
The crossover data was obtained from shipping records. Problem is, those records would be a near perfect unblind of the trial.
True, NWBO was never unblinded. Just the shipping records under control of LP were used to generate unblinded summaries to LL et al.
Does LP have a Chinese Firewall between the two sides of her brain?