Could I get some more detail here? Which /sorts/ of OXPs lead to the biggest performance hits? Missions, stations, ships, equipment, weapons, HUDs, retextures, ambience, others...?
That's a bit tricky to answer. It really depends on what an OXP exactly does. For example, if an OXP is switching huds on every frame, chances are that performance will get a good hit. OXPs that contain heavy on timers and callback functions scripts, when not optimized, might be a performance hit. The problem of course in this case would be to identify those heavily scripted OXPs that are not too optimized. Another possible performance buster might be OXPs that rely heavily on shaders, when your gfx card is low end. Finally, remember that OXPs, being community products, are not quality-tested by anyone other than their creators. Sometimes OXPs simply cause severe lag due to bugs (there was recently a case of na OXP that was basically filling up the universe with entities, without ever removing any, which resulted in single-digit FPS followed by an unceremonious exception in the core game within about 25 minutes of gameplay).
An easy start would be to remove the OXPs that are often reported in the log as "having run for x.xxx seconds and have been terminated". But again, this is not gospel because the Developer Release of the game has very strict limits imposed for scripts, while the end-user version is more relaxed. So, even if an OXP is reported as running for too long in the Dev version it does not necessarily mean that it is the main cause for eventual performance drops.
Moral of the story: Decide what you want your game to be like. Then load up OXPs that do just that and do that one at a time so that you can verify that all runs well after each addition.