A few months ago, following a lead by BrowserUk and using source code (slightly rearranged) by blokhead, I started collecting Selected Best Nodes, until I had 2000 of them.
Once I was reasonably confident that I had the first 2000 best nodes of all times, I started playing with the resulting database, and eventually I made a simple search engine, from where I can get some statistics by section or author, and lists of nodes, customized by section and rep.
A few examples:
Stats by section, sorted by number of nodes: (but you can get stats by author, or sorted in different ways)
List of Meditations and Cool Uses for Perl nodes with rep of 200 or more: (but you can list by any other section or rep value)
More than my description, I believe that using it directly would be easier to understand. So have a look at http://gmax.oltrelinux.com/cgi-bin/cbnodes.cgi, where you can also download the current database in SQLite format.
I will appreciate comments and tips for enhancements. Actually, if the gods decide to implement something like this in the Monastery, I would be very happy.
Thanks to demerphq, BrowserUk, dragonchild, and blokhead for providing useful feedback and tips.
Modified links in example table. Since it was cut-and-paste code from an external page, its links could log out the readers. Fixed now. Thanks to talexb.