If you've only got 100 pages, maybe you'll get a 70/30 rule. And as you get quite a bit bigger, the statistics will very likely push you more towards a 90/10 rule. But the basic Zipf curve will be there for virtually every site that isn't tiny. You simply can't multiply an increased page count by your current earnings to predict the future. efa says:
First, let me say I'm highly confident in the Zipf distribution. Whether that leads you to precisely an 80/20 rule is not important, the important thing is most of your pages will earn you less traffic and revenue than your per-page average.
If you've only got 100 pages, maybe you'll get a 70/30 rule. And as you get quite a bit bigger, the statistics will very likely push you more towards a 90/10 rule. But the basic Zipf curve will be there for virtually every site that isn't tiny. You simply can't multiply an increased page count by your current earnings to predict the future.
This is an interesting point for which I think there is less broad data to go on. We would like to hope that the Long Tail will mean that, even though those poor-paying pages are paying well below average, that they still sum to a large amount.
Personally, I'm growing increasingly uneasy that it's possible that, with the huge explosion in competition for Google SERPs (and I'm focusing solely on free SE traffic scheme sites here), it's possible that the Long Tail cannot be easily maintained.
In my own pages, I see an increasing number of "dead" pages (absolutely no traffic/revenue) as my page count goes up. I don't have enough numbers over enough time to say what the trend is for me, and I don't know of any research that would give a peek at what the overall trend is for many webmasters.
I see my Long Tail is getting stomped on. It does not extend out to the right anywhere near as far as I would have hoped. Or, maybe it's just that my Long Tail is too flattened: if a page only makes me $.50/year, that's indistinguishable from zero as far as I'm concerned. I don't have 10,000 such pages to make the revenue from a too-flattened Long Tail multiply into something worth noticing; nor can I afford to write new content for wages of $.50/page/year.
So yes, that other 80% is making you some money. Makes you money even when you go on vacation. It's great. But I'm becoming much more sensitive to the fact that it's increasingly easy to create pages that, perhaps after an initial burst of free SE love, provide absolutely 0 traffic/revenue. They still serve other purposes, of course, (another place for already-arrived visitors to hit, intra-website page rank and SEO, etc.), but I do not generally want to spend time creating such pages.
In some ways, I think the Zipf curve is the central problem of being an AdSense publisher (at least the Good Content + Free SE Traffic + Qualified Visitors kind of publisher). How can you minimize the number of pages that won't pay you well for your time (even if multiplied times 10 years of earnings)? How can you raise the percentage of pages you write that will, over some plausible number of years at least, reimburse you for the time you invested? How can you squeeze a bit more revenue out of that flattened Long Tail to turn it from Waste-Of-Time into Paid-For-Itself?
I would sure like to see some data drawn from a large pool of AdSense publishers on just what their Long Tail looks like (Google has the data, but I somehow doubt they will share.). By the time you have even just 1,000 pages, a small change in the height of that Long Tail can mean a big difference in your overall results.