I have a love/hate relationship with the S3 dataset in the Rules Engine.  I spend a lot of time trying to get files formatted correctly so that the rule picks up the data correctly and pulls it in for my rules to run. 

1. Would love if the mapping process was a little smarter - i.e. could identify the formats of dates, date/times, etc.
2. Would appreciate if the errors were a little more clear.  My latest example was due to an incorrect date format, but that date was in several columns so it was difficult to figure out which was the issue.