Should BiBiGrid try to guess what flavor/image is meant if no match is found?

Currently if a flavor/image gets rejected an error message appears and the program terminates. Using difflib or a similar package comparing the failed flavor/image to feasible flavors/images and picking the best match is possible. That would allow BiBiGrid to rescue and create the cluster. Such a rescue would be logged as a warning and shown to the user. Maybe the user has to explicitly accept the new suggested flavor/image.

However, this comes at the price that maybe a wrong image or wrong flavor is selected.

Is this kind of rescuing desirable?