The project is fairly open-sourced to a degree to check it out prior participating, and it indicates around ~160 submissions by participants already.
Though, the question some may imagine: would this research results be included into the phishing models training eventually, and enhance the phishing itself in turn?