I don't think at this point we can be sure that there is a connection between PUFAs and oxLDL. Kismet over at Imminst addressed this issue in a post:http://www.longecity.org/forum/topic/48670-the-oxldlox-stress-pufa-connection-is-very-dubious-evidence/
The issue is that PUFAs could increase oxLDL in the serum because since it is lowering LDL oxLDL particles are being removed from the intimal space. However as I wrote about in my recent blog post (http://hanswuhealth.blogspot.com/2011/11/paleo-folk-beware-of-your-cholesterol.html) it is quite possible that oxLDL may be the instigator, however from then on there the process moves on at a pace determined by many factors (less cases like FH).
If we are arguing on the basis of PUFAs are less stable, thus they COULD cause oxidation in vivo, we don't really know that to be true. By that reasoning, SAFA would be the most stable, however studies comparing the two (albeit limited in number) show that SAFA leads to more oxLDL (possibly inflammatory? discussed here: http://hanswuhealth.blogspot.com/2011/11/lipid-profile-and-goals-part-ii-mufa-vs.html).
While there is lack of data showing the benefit of no PUFAs, there is also lack of data for increased PUFAs, for now I think we can settle down for 20g total a day. 2:1 ratio of w-6:w-3.