[edited from support email]
MediaLab gives a warning when there are duplciate variable names within the same questionnaire, BUT not within the same experiment. Wouldn't that be easy to implement?