Some dynamic URLs get spidered while others do not. It seems to depend on their complexity. Something as simple as your expample tends to spidered. More complex URLs with with session IDs etc have a harder time.
I think we should be able to. Now if you're using "source=" to represent the location that the user came from, that's pretty inefficient. Then the same page can get indexed multiple times with a different url. If that's what you're doing, I'd drop the source param and get the same info from your logs (check for the referer field).
But if you're just talking about dynamic urls, we can crawl most of those. As someone else said, it's best to avoid totally weird/long params though.