It's not exactly A/B testing but I've stuck in a rand() function so that the Today page randomly shows either the actual time of last update or the (approximate) time since last update.

I'm not sure which I prefer.