i think you missed that most of the development activities in the pd world are based on voluntariness and collaboration. That's why noone can actually expect pure-data.org to be accurate or fit one's needs.
i think there would be, at least, a moderator or even better, an integrator which actually tests submitted software. this is obviously not the case and the success of PD really depends on this. i do respect Dave Phillips' work to that extent, he's a really useful guidance for making sound on Linux.
You have read (and maybe laughed) about the efforts towards a unified library or RPM packages. I find it a pity that this is a lot harder if
there
is no common resource for pd stuff.
it's even worse when the common resource spreads inacurrate information and apply an "anything in, anything out" policy.
yves.