I'm really much more interested in what the two browsers do when you have 25+ tabs open at one time. That's the case that's likely to have consequences for system-wide memory usage. Even more than that, I'm interested in the case where you've had the browser open for 8 days and have opened and closed hundreds of tabs and thousands of pages.
Having separate processes buys the ability to eventually close a process (when all the tabs/pages associated with it are closed) and reclaim all the resources. This is handy if there are memory leaks (with proper programming, there shouldn't be, but there *certainly* have been in Firefox -- a browser is a pretty complex piece of software). It also seems like it would be helpful for memory fragmentation: if you keep the entire browser state separated into N buckets (processes' memory spaces), then when you free one bucket, you can free every single page in it. You never find yourself in a situation where you can't free a page because some unrelated thing (that happened to call the same allocator you did) has 100 bytes of data still occupying that page.
You're right that it could lower the consequences for sloppy programming, but I don't care that much. Anti-lock brakes on my car lower the consequences for sloppy driving, but I still want them.