Outlook and Outlook Express are two completely different products. I'm fairly certain I can write a CSV parser without createing 'tons of code.' More importantly, that's not the correct trade-off since there are numerous APIs in windows for abstracting access to various types of text files (csv, xml, ...). Thus if you do it right, you wouldn't be writing any parsing code at all.
If this whole discussion is truly over 15 minutes of work, then why are we discussing it? On the other hand, if the 15 minute estimate becomes hours or days when the full capability of the feature is implemented, debugged, and tested, then priorities become an issue again.
You are right, it is better to 'design' features in early. Implementation is another thing all together. During solution design, it's best to start with generic interfaces when possible and only move to proprietary solutions (outlook direct) when other factors (schedule, budget, specific customer requirements) dictate.
That's my 2 cents on this issue.