As some in the comments pointed out, low-HD monsters are vulnerable to mass attacks like sleep spells and fireballs, while high-HD monsters like good old tyrannosaur is unaffected or, at least, more resistant. This is an important point.
Also, I wanted to simulate a couple of other fights with creatures of higher HD. So I ran six quick melees between five ogres and a tyrannosaur, both of which would be worth 2000 experience points under the proposed system. The dinosaur won all six fights rather easily, as it almost always hit and nearly always killed one ogre each round. It seemed to me like the ogres rolled really badly on their to-hits, which didn’t help them in a fight where the odds were already stacked against them. None of the six fights were close.
I tried another round of battles, this time between ten gnolls and one tyrannosaur. This was a lot closer, with the tyrannosaur winning three of five battles. In one of the battles, it finished off the last gnoll with only 7 hit points left.
A few commenters have pointed out that the majority of XP should probably come from treasure in most standard campaigns, and I agree. In my current games, probably around 75% or more of XP comes from treasure. I wouldn’t want this to change by very much, if at all. I don’t know that the fact that most experience comes from treasure should affect the awarding of monster XP one way or the other, though.
Current XP calculation methods do not take a monster’s treasure rating into effect, so whether or not a monster is likely to have a lot of treasure does not affect how many XP that monster is worth. If monster XP creeps up a bit, the game master would have to keep an eye on things and keep total XP awards and the combat:treasure ratio where he or she wants it, but that’s something that game masters already have to do.
Finally, I want to be clear that I’m not championing this idea. I don’t even use it currently. However, I do wonder if it might not be an easy way to award XP while maybe doing a better job modeling the danger of groups of weaker monsters. It would probably make advancement out of the first few levels a little bit quicker and advancement into the higher levels a little bit slower, but I don’t see that as a problem at all. I personally think the best adventuring usually happens in levels 4-8 anyway, so if this system extends the period that PCs are in that golden range, so much the better.