When was it smoking? Was it smoking at Idle, or when the RPM and oil pressure was higher? I ask this because the oil delivery flow increases with pressure, and there is a possible point where the drain / return line capacity could be overwhelmed. When this happens the oil is forced out instead of draining. This will cause some of the oil to get past the shaft seals even if they are perfect. I have had this on my own stage 2 cars as well as customer's cars. I believe the higher pressure oil, and therefore flow, that is supplied with a turbo saver type filter and a fresh tight build with hi cap pump is greater than required for the flow rate to feed the turbo. There several options to deal with this. The drain to the pan was not a very good one for several reasons. The drain to the front cover straight under the turbo with -10 line was not bad, but it requires the removal of the front cover and all of the stuff that is in the way, (big job), even with FMIC. I addressed this on one customer's car with an inline adjustable hydraulic bleeder valve like we sometimes use for boost control for integral gates. I plumbed in a fuel rail pressure ga between the bleeder (adjustable restriction) and the turbo inlet oil fitting for an idea of the oil pressure that it was getting. You could use an electrical sender and a ga inside the cockpit if you wanted. The idea was that the adjustability would allow a precise restriction for the specific application without having to experiment with changing out set orifice restrictors. BTW, John Craig warned me not to do this. He said restricting the oil to the turbo was not the way to do it. He said to address the issue with increasing the oil drain capacity.