Dr. Cutlip said that while thrombosis risk with Cypher increases with time, there was also late risk with bare metal stents.
The published data, such as it is, tends to bear this out. The overall mortality for DES vs BMS shows that, if anything, the mortality is higher for BMS. (the problem is that BMS vs DES trials were powered for MACE, not survival, so the number of deaths is small and thus does not reach stat sig - but the general trend appears to be that DES is equal to or better in long term survival)
Personally I couldn't care less whether I croak from an in-stent thrombosis or restenosis. I only care about the odds.