Regarding #3: you could use a Schema reader to read the csv schema and then compare it to what it should be. If it matches, you can use a FeatureReader to read the actual data.

It may also be worth exploring the workflow before the data gets to FME, if there's case differences being introduced there that's obviously not a good thing in terms of consistency.

@dutchtower​ There's a webinar on schema drift that might give you some ideas - including using the SchemaScanner or FeatureReader to get the schema and compare it to an old one.

But... when you add a CSV reader to your workspace, under the reader parameters, Advanced, Field Names, try setting Case Sensitive Field Names: No

I added a new CSV reader and set the Case Sensitive Field Names to No and replaced the original CSV Reader and SHAZAAM! it worked. I'll take a look at the webinar Mark mentioned and see about adding a Schema drift alert. Thanks, everyone!