There is definitely no benefit to leaving the default /productcart/ sub-directory named as is. There is also no benefit to renaming it to something like /store/ or /shop/ other than purely cosmetic. There is also about nil SEO benefit to naming it something more key-word specific as it's simply plastered all over the entire store and therefore becomes basically irrelevant (bordering on "SPAM"). Personally, I think the whole idea that it has any SEO relevance is merely an excuse for rationalizing their existence in the first place. It's just as probable that the double sub-directory structure is watering down relevance (as, for example, Google will interpret this as less relevant because it is structurally further removed from the root). There is plenty of debate on whether this is true really, and sadly Matt Cutts hasn't answered my email on this yet ;-)

Personally, I believe that the best possible thing to do is to rewrite them away entirely, as if they didn't exist. This can be done using the IIS 7+ URL Rewrite snap in if your host provides it, or on older servers if they are running ISAPI_Rewrite. Here are a couple of examples:

http://www.nancyboy.com/Skincare-c20.htm" rel="nofollow - http://www.nancyboy.com/Skincare-c20.htm

http://www.hhandc.co.uk/Alsace-c105.htm" rel="nofollow - http://www.hhandc.co.uk/Alsace-c105.htm

If your host provides one of these URL rewriting methods, PM me with your email address, let me know which method, and I'll shoot you the rules to do this.

My understanding of what Cutts is addressing in the video there is "on-page" keyword density (things like title tag, heading tags, page content such as paragraphs, text in anchors, titles in anchors, alt tags on images, etc.). That is, the minimal keyword value that comes from URLs is worth considering, but it's small potatoes compared to these other on-page factors above.

So, let's say my ProductCart product page is about red widgets (to use Cutts's example). Either by default (if I named my product Red Widget), or because I supplied a custom title tag for this product ("Red Widgets"), that title tag will carry the most weight of anything on the page. If I'm using PC's "Keyword-Rich URLs" features, then that bit in the URL will have a tiny bit of relevance -- all for the search term "red widgets". If my /productcart/ sub-directory were named /red-widgets/ then we'd get into that plateau or point of diminishing returns Cutts describes. If it were named something else that isn't semantically closely related to red widgets, then the name of the sub-directory in the URL is basically irrelevant.

Now, if we translate some of that on-page theory to the global site level, it's pretty easy to see (esp. now that domain names don't pull weight here anymore as you point out) that the keyword value possibly derived from the name of this sub-directory is about nil (on that plateau Cutts describes, though I don't expect this would tip into diminishing returns). If every page in my store has a URL that includes /red-widgest/, then that basically it washes out as irrelevant for a search. If every title tag had the same phrase, then pages would start to be demoted. The latter is SPAM, the former is structural (so not so likely to be considered SPAM).

But here's the rub, IMHO: it's really not about SEO at all; it's about application structure.

Let's compare ProductCart with WordPress for an example.

WordPress, by default, also natively installs into a subdirectory called, well, /wordpress/. In both cases, WP and PC, this is because one might have other stuff in their webroot and wouldn't want that interfered with when they come along later and install one of these applications. Fair enough. Good architecture (however, it's simple to install WP in the web root, but not so much with PC -- though I expect I could do it without too much effort). 

From there we get into the sub-sub-directories. WP also has an admin sub-dir and an includes sub-dir just like PC. They are pretty much theoretically equivalent here (though they use very different methods for using them, when it comes to the includes). The big difference is in PC's /pc/ sub-directory, which is basically where the public-facing application is housed and that cannot be changed without basically rewriting the application. 

I've never seen anyone in the WP universe claim that one should install WP in one's web root with some keyword phase to replace the /wordpress/ directory -- either install the application in the web root or rename it to /blog/ or something with meaning to the user (PC's equivalent might be "/store/") on this level. No one tries to rationalize this application structure by claiming it has any SEO value at all, yet this seems to persist in the PC community -- I expect precisely just to rationalize the application structure (esp. because there's not getting rid of that /pc/ directory!).

So, let's just say I've an .ASP site with a bunch of stuff in the web root. It would be wise to then install PC as something like /store/ or maybe even /red-widgets/ -- just don't rationalize that as beneficial for SEO. I would probably also install WP as something like /blog/ or maybe /red-widgets-digest/ -- and again, I wouldn't rationalize that as beneficial for SEO. This is just clean architecture.

Finally, if I don't have anything going on in my web root and am installing PC or WP, then for the latter, I'm just gonna install WP in the web root. With PC, I'd rather just apply rewrite rules to rewrite away the /productcart/pc/ sub-directories to get the same effect. Either way, I'm back to SEO issues that are not in debate.

I'm with Brett: let's here what folks have tried and tested here and what solid SEO theory applies here (such as Cutts's vid). 

The absolute bottom line I think everyone will definitely agree with is that one should rename their /productcart/ sub-directory to something more relevant to their site -- say "/store/", e.g. And sure, if you are going to keep that architecture, then why not "/red-widget-store/" or something. Probably won't hurt anything. Does that actually have any SEO value? I think not. Would it be better to rewrite the /productcart/pc/ directories away as if they didn't exist when the rules for this are so easily applied (not that they are so easy to create from scratch)? I'd tend to error in that direction so that I have a tighter set of knowns in my SEO data and testing.

If someone has a contemporary (post Panda update) set of data that seriously suggests keyword relevance in SERPs coming from the way they named their /productcart/ sub-directory, I'm sure many of us would be all eyes on it. Else, short of that, maybe we should stick to what's solid and abandon the creed to take on faith that this directory structure has SEO significance and rather just admit that it's just structural and consider the (probably minor for most) differences between retaining that structure or rewriting it away easily (to get it out of the way, as it were).