per discussion with @dideler, python-feedparser provides functionality for url lookup and fetching of the feed which includes error handling and is well tested and well-maintained, if we replace this we need to implement and test this ourselves so it may be worth looking at including python-feedparser source code in the lib folder (possibly as a git submodule or some other approach. It is MIT licensed so is compatible with our project). This would reduce the end user's need to download an additional dependency to run the software.
Alternatively after some research I believe we should be able to implement the feed lookup and fetching using QUrl class which is also in QtCore.
If you can do it all in Qt I'd say go that route. QUrl should cover what you need here but you'll need to do more research to make sure.
The key is to ensure that your new tool works exactly like the old tool that it's replacing. One thing of note is the old tool despite being called the "RSS Import tool" it's not actually very generic. It was designed specifically for fosslc.org's RSS feed. Example feeds we've used in the past:
I think every site my be different. I actually envision a rewrite of the RSS tool would include turning it into a full fledged Freeseer plugin type related to bug #151. This way other conferences and sites could make a RSS import plugin that is specific to their conference RSS data structure.