Look, these are worthy causes you refer to, nobody disputes that. but you're basically presenting a a kind of zero-sum fallacy - there is only so much labor power available, disease and poverty are humanity's most pressing problems, therefore all other activity should halt while those are tackled.
This is a naive viewpoint. I am not going to argue that Facebook is somehow better than a cure for malaria, but the problem facebook solves - 'how can I efficiently keep in touch with the many different people I know?' - is important enough for individuals to find it very useful, and the network effects, plus what we learn about large-scale social networks, may well turn out to have broader implications. For example, a recent CDC study drew the conclusion that the most effective way to handle the current swine flu problem might be to target vaccinations towards young (Youtube seems even more flippant than facebook, devoted as it is to sharing video for entertainment purposes. On the other hand, if I wish to learn about malaria or some other serious problem, YouTube offers a way to access video about subject almost immediately. Putting such stuff on TV reaches a wider audience than any given YouTube video, but by necessity a passive audience and only at a given moment in time, rather than when someone is motivated and active enough to seek out that information. Again, not the goal of YouTube, but such a useful fringe benefit that we now take online video availability as as essential factor in future internet growth...and indeed, MIT are using YT as part of their strategy for OCW. One person's luxury can easily turn out to be someone else's utility.
Frankly, I think HN is quite good about propagating news of innovative non-computer hacks as well. Dean Kamen's water purification system was recently the subject of extensive discussion, and the criteria for what constitutes hacking is already pretty loose - tech-related, cheap and innovative generally qualify.