That makes a lot more sense then. It's still a misleading statement to say "writes" vs "transactions" since you could (potentially) make fewer writes and support more transactions. The ratio between the two is a measure of efficiency, but only transactions matter to end-users.
You're right that one number trades off another. I'm not sure that only transactions matters, though.
Tracking logical writes makes the test comparable to tests of key-value stores that can only update one key at time, which is pretty much every other distributed database.
