I think there's something worth looking into here. Great insight about the "crap filter". What would you say is the most technically ambitious part about this project? I'd reckon that it is the resources to constantly pull updated content from all monitored pages at an acceptable frequency e.g. daily/hourly/minute-ly.