That's right. Brandt has suggested, in one of his numerous on-line discussions of this, that an easy way to see which spokes see significant tension change when a bicycle wheel is loaded is to pluck the spokes both loaded and unloaded and see which spokes have a discernible change in pitch. The bottom spokes do, the top spokes do not.

Here is an interesting analysis the results of which agree with Brandt: http://www.astounding.org.uk/ian/wheel/

Of particular interest is the drawing showing deviation from roundness; greatly exaggerated, of course, because there really isn't much deviation from roundness.

Of all the things that have been written here the worst is, "[t]he bottom spokes do nothing to resist vertical load on the hub." That's beyond wrong. It would be nice to know what buildings this person has designed so I could avoid them.