We've previously lamented and explained some shared hosting performance issues, and I'm happy to say that various countermeasures we've implemented over the past 60 days have helped a lot. The combination of hardware upgrades, software configuration changes, and the cooperation of dozens of understanding customers has improved performance a great deal. The graph below is indicative of the storage system's 'load average' we've experienced over the past 12 months, and as you'll notice, things are much happier now.
Also, we're in the midst of a storage architecture "re-engineering" that will make a more dramatic impact on our storage scalability, as well as opening the possibility of some interesting new features. Stay tuned for more updates.