I just dug out my copy of "Lost Worlds...." and re-read some of the discarded chapters and I think I'm seeing a hint at something here.
In one passage, Clarke talks a lot about how most animals across the cosmos rely strictly on instinct. A rare few stumble onto true intelligence, sentience -- self-awareness. For most of them, they shake off the haze of instinct, see their own looming death and the spark immediately goes out. That's why the alien race decides to give these bright apes a little nudge. So the spark of self-awareness catches fire in Moon-Watcher (I remembered the name wrong earlier) and what is the first thing he does.... He kills.
Now think about HAL. He's a computer -- following a program -- just like an animal operating on instinct. But now he's become sentient. Become more than the sum of his programming. And so what does he do? He sees his own looming death and he kills. Just like Moon-Watcher did.
I think that is why Clarke chose to "save" HAL at the end of 2010 by having him merge into the Monolith's consciousness as Bowman had. HAL had made the same "leap" as Moon-Watcher did so from the alien's point of view he was just as worthy.
Here's a very interesting presentation of what someone else thinks Kubrick was up to.
http://www.kubrick2001.com/
