As the years go by and my new content continually pushes older content into the bowels of my internal link directory I'm finding that search engines, Google in particular, are changing in how they handle it.

PRIOR TO PANDA:

Before Panda rolled out I found that linking to previously top ranked content from newer content kept those old great posts from slipping to page 3 in serps and beyond. When enough of these formerly heavy hitting articles are pushed deep enough into the archives I found that creating a new archive for them often worked well. The best content could be kept within 3 clicks of the index and all was well.

AFTER PANDA: (until a couple months ago)

When Panda rolled out I found that if my site attempted to maintain too high a percentage of lowly ranked pages then the whole suffered and when I slowly began to cull older content, especially if it had any sort of affiliate offer worked into the article, sitewide rankings improved. note: I used 410 error pages instead of 301/302 redirects, and I removed all links pointing to these pages. Less(pages) became more(traffic).

RECENTLY:

Now I'm finding that a hybrid approach is working best in which removing only the worst of the worst pages from my site (e.g.profile pages, pages with little content because the subject dictates not much can be written, news stories found talked about on most major 'social' websites etc) is enough, with a twist.

Deleting only the worst offenders and then moving the old awesome articles right off the site onto a NEW domain with a 301 redirect is yielding the highest level of traffic overall. The old content is indeed appreciated by search engines once again, when given a breath of life that page 117 of my directory hierarchy could not provide. Simply linking to it from newer similar content was not as effective as it once was.

I now call this repurposing the content, it works in small batches at a time, but I'm wondering how long it will be before this becomes innefective too, and I'm wondering if it's safe to do out of fear it's going to cause major repercussions to my original site in some future algo change.

Note: care is taken to choose the 'right' pages to move, all related with similar backlink profiles and traffic 'value' as reported by GWT.

Since increasing web traffic is the goal, and both culling and repurposing seem to help achieve these goals right now, is it a good idea to continue culling old content, repurposing it, both... or neither? What problems do you see with the practice, besides owning a growing number of domains instead of the traditional growing number of archive pages?

Also, on a different note, does this open the door to buying and/or selling existing content without worry if proper 301's are used? It would seem that you, given enough money, could indeed now go and pick the best pages online for any given keyword and benefit from bringing them together under one domain (even if that was not a search engines goal).

On yet another different note, my competitor has written some AMAZING articles on tax implications within this particular subject but the subject itself is not about tax and these articles are lowly ranked for tax related keywords as a result. If he creates a tax website these are the kinds of articles one would hope to find. Since he cannot shake subject A to improve article results for subject B on the one website does it not make sense to create site B all together? From a web perspective, if this is indeed rewarded by search, I fear that the number of domains online may grow exponentially and it will become harder to rank in general as a result.