I would call it mostly harmless. As long as you are still going by the content and quality of the posts, and not by who wrote them, how you generate your list of posts to possibly rate should not matter. However, I might also label this as a little on the "bad practice" camp, as you will probably have some natural bias and will not evaluate the posts the same as if you had just wandered across them, no matter how hard you try. IMO, the author of the report should never affect your voting - vote the post, not the author.

If you think about it, you are not introducing any negative effect this way. You can only vote for a post once, regardless of its age, therefor if you find an older post that deserves a ++, by all means vote for it! (And likewise, if you find bad posts, vote them down!) The real bias here would be your choosing to abstain on a node because you want the author to be promoted despite the crummy post. However, as most people here seem to vote ++ or not at all, the bias is negligable.(Compare the rankings of the Best Nodes with those of the Worst Nodes. There is a substantial difference in absolute value.)

So happy promoting.

Sounds to me like you could be biasing on the top posters and not giving the newbies a proper chance.

Or alternately, most of your votes would be spent on the newbies because the have steps that are closer together and there are lot more of em...

Now I'm confused. Hmmm, well if you can't read all the posts in a day, it can't be that bad of a practice... =)

--
$you = new YOU;
honk() if $you->love(perl)

This is cool, although I shun the point system (now I've got my photo on my own node :-)Unless there is an overriding decision to the contrary, I will adopt this.

--