What blind spots exist around the trajectory of quantum machine learning scales beyond early adopters, and what uncertainties remain unresolved?