This is due to:
1.) My indiscriminate reading habits. By this I mean both my tendency to read a lot of different things online, but also my slapdash way of doing so. There's a lot of ALT + TABbing that goes on in my day, and I can often become confused even while reading something about where I am and how I got there.
2.) My use of an RSS reader (or whatever you want to call Bloglines). I remember reading something that accompanied an exhibit of illuminated manuscripts at the Getty museum that explained that the drawings along the margins were sometimes used to illustrate the text, but more often were used as mnemonic devices. You would read something and later recall that it was on the page with the picture of the rabbit in the lower left-hand corner, or the pertinent passage had a bird swooping past it on the right margin. Blogs and websites have a similar way of working by making a sort of subconscious imprint of their layout and color scheme in my mind, so that I can sometimes recall an annoying pop-up ad or flash animation to lead me back to where I found it. But because the feeds all look alike, I remember that I read it in Bloglines, which means that it was on a blog, but which blog is sometimes lost completely.
3.) Information and ideas continue to percolate long after I first read them. I rarely have a fully-formed response to something when I first read it. I usually turn it over in my head for a while as I go about reading other things, and those things add to the mix and make me think of relations between topic A and topic B. I ponder these things as I work, eat, watch TV, listen to the radio, converse, drive around, and generally live my life, and by the end of the day it's turned into something completely different, but it's still based upon what I read earlier that day, earlier that week, or even sometime last year.
The added problem of the unreliability of one's own memory is the unreliability of the web itself. I spent time this past weekend hunting down online versions of articles that I had printed or clipped over the past several years. In some cases, I had a URL printed at the bottom of the page that was still good. In others, the URL led to a pay-only archive, or told me that the address was no longer current, or in a few cases led to a totally defunct site with a domain name up for sale. Where I didn't have a URL to start from, I had to Google a combination of title and author and hope for the best. Sometimes I found reprints on other sites, or the author had put the piece up on their own site. But for a significant number, I was completely out of luck.
So what if you find something on the web, diligently note it and bookmark it, but when you go back to find it again, it's gone? A book can go out of print, but copies still exist. It may not be easy to find, but it can be found. Where do web archives go when no one wants to maintain them? What to do when a mass of articles exist solely on a hard drive somewhere?