I'll respond to your question with others :).

If you have a website, how would you determine that a particular visitor is a webmaster?

How would you determine if s/he's visiting professionally or as an ordinary surfer?

are you able to determine what percentage of your ODP users/visitors are a, webmasters and b, general web users

I think the short answer is this is "No"

(...)I think the answer would be illuminating for all of us

I completely agree, but as jimnoble points out, even if we had access to the webserver log files for the front-end server(s) of dmoz.org (which editors don't - that kind of stuff is for ODP staff/Operations only), how would we tell the difference between webmasters and 'normal people' ;-)?

The only useful comparison I can come up with is that Alexa ranks dmoz.org at 204 and WebmasterWorld at 230. Surprisingly close.

Given the kind of people who might install the Alexa toolbar, and are therefore skewing the sample, I would put my money on BOTH those rankings being a bit high...

...but if we got Brett really drunk and persuaded him to spill his traffic/vistor figures for WebmasterWorld we might end up with a reasonable first guess at the dmoz.org traffic/visitor figures :-)

I fail to see the relevance, but aside that, there are definitions needed:

Define webmaster:

How much content do you need to have online before you are a webmaster? Is having a "under construction" page online enough to make you a webmaster? Is a person having a homepage with a resume on it a webmaster? Is a person having made a large web site for his employer a web master even if the tas is done long ago and handed over for maintenance to others? ...

Or any other criteria you need?

Define relevancy:

Suppose I have a leading website on widgets, but today I'm looking to buy a foobar, is it relevant that I'm a widgets webmaster?

Define general web users:

Is a spider a general web user?

Define how to measure:

How would the ODP know you are a webmaster?

How do you know any of your visitors are webmasters? (If you'd know you might want to change all your ads to "please link to me" banners)

How do you handle proxied requests?

How is any single entity going to know the exposre of all of the direct and indirect uses of the ODP data?

Define what's using/visiting ODP data:
With the rdp data available to e.g. search engines, to websites to include as their links section, to mirrors of the ODP how much ODP data needs to be used in what way before it's considered a use?

That said, I think it's impossible to answer. But suppose you get an idea through a sample of the population based on clear definitions, how do you think it's going to be used to enlighten us all?

Just to complicate things, many ODP users don't visit the site at all. They use the copies at Netscape or Google or Alexa.

I think that, whatever answer you come up with, it'll be like "They say the Princess Infanta's eyes are as blue as the Star of India."

That is, someone you don't know is rumored (with what reliability you don't know) to have said (whether truthfully, or on any valid basis, you don't know) that something (which you've never seen) is as blue as something else (which you've never seen).

My personal reaction is to try to dissuade you from desiring such an incredible illusion of enlightenment.

try to dissuade you(...)

veering slightly off-topic, but hutcheson is the one to ask...

Has there ever been any significant (public) discussion of the hardware and/or software that dmoz.org runs on, or is this kept under wraps?

I'm looking for more detail than the stuff rpfuller posted a while ago ;-)

Good sirs, some times enlightenment comes with an absolute flavour

Thus, if the Captains of Dmoz where to find it beneficial to have more intimate knowledge of their "users", they might decide to

bait the directory with various savoury "webmaster bait" like automated login accounts, where by the logged in webmaster who had properly verified his/her identity would be regaled with tales of the status of their submissions, rebuked mightly for their transgressions, lacking their customary anonymity, most might behave more decorously to the ever so civil odp editors,

and oh yes, such webmasters would recieve a long lived cookie, that they would treasure an never let adaware delete

vite, you can dream. But it's like trying to imagine how a 1957 deSoto could be used to generate radioactive isotopes (really very simple, once you replace the motor with an atomic pile). The processes, databases, and so on would have to be too different from what they are now, for information to be either available, publishable, or meaningful -- let alone all three!

And that's without even discussing the adversarial relationship between the "unique content" mission of the directory, and the source of 90% of site suggestions.

I can't imagine the ODP techies seriously considering implementing anything like this in the foreseeable future.