<pre-flame preface>

This is meant in a good spirit, so no real flames back please.
< />
I have been at this software development thing since 1976. I won't bore you with war stories of the old days though. -g I have dealt with supporting user-extensibility issues on a number of systems, some far larger than vBulletin. Nuf' said.
Professionally speaking, the hack installation conventions we use, well, they suck. One can so easily end up with so much tape, wire, and safety-pins holding the thing together (after installed a bunch of hacks I mean), that the thought of upgrading to a new release sends chills up the admin spine.
Bottom line: Hacks should be supported by an SDK so that, as far as possible, they can survive an upgrade of the base vBulletin. For some types of hacks, this would not be possible, but for most it would be possible. This would require a hack support module in vBulletin, but once such a module were installed, supporting the installation and maintenance of hacks would be much easier.
ps: Even the name hack is wrong imho. The usage of this word implies "not professional" or "not a proper modification".
pps: HTL, although a very fine piece of work, is not the answer.
Cheers, CarCdr