> Verily cited insufficient consistency in the correlation between tear glucose and blood glucose

For a research project, they should have measured that in week one, and canned the project just 3 weeks in after a bunch of test results on swabs from peoples eyes couldn't be closely correlated with blood glucose after controlling for temperature and humidity in an excel spreadsheet.

How did it take years for them to figure this out? Smells of bad management to me - the engineers saying 'It's so nearly working!!' for years just to keep their jobs, and management believing them.