I found some AWESOME info about smart objects on Wendyzine's blog. She explains how a smart object itself isn't "editable" but for good reason: when you resize a smart object, Photoshop references the original size when shrinking/expanding the layer. So, you can place an element on your page, turn it into a smart object, and then resize it up or down while you play around with clustering on your page--all without worrying about losing image quality. :) You'll notice that if you resize a smart object to 75% its original size, the next time you resize it, the Width and Height percentages won't be "100": they'll be set to 75%, in reference to the percentage that you originally shrunk it. Why does that matter? Because you lose image quality each time you resize a layer (especially if you resize UP after resizing down). So this way, your images will stay crisp and clear (or rather, as "crisp and clear" as the original was!).
This also works really well when you resize for the web (designers, take note: my previews look gorgeous now that I turn them into smart objects before resizing for store previews/"folder"-size previews!). I've noticed a marked increase in image quality. I don't have to use a "smart sharpen" filter on the layout after shrinking my 12x12in and 300dpi layouts to 600x600, 72dpi because the image is already super sharp (note: I do still use the "bicubic sharper" option when resizing).
Anyway, here's a photo Wendy shares to show you what I mean:
HERE to read the full tut.