In case anyone else is as completely lost without context as I was, the NoSQL meetup was apparently a group of people interested in key-value stores, and Dynomite is the author's clone of Dynamo, a key-value store from Amazon.
Good point, I completely forgot that not everyone is as knee deep in distributed datastores as I am. Thanks.
I'd like to point out a single summary for this: "Let's actually act like Computer Scientists and run controlled experiments."
The "Scientist" part means something more than "not an engineer."
Still, it's nice to see some headway in bringing some honesty to benchmarks!