The idea is interesting indeed - and easily doable as well: Entry.all gives you all the links on the front page, you'd just have fetch the main page between regular intervals of time (say, once every hour) and then save the entry information to a database.
Could you elaborate on this? If you can provide better "specifications" I'll implement this and add it to the library.