I thought Lunardi has gotten every team correct every year except for last year where he just missed with SMU.
He gets nearly all of the teams (sometimes all), but his seeding is generally off. I'm not saying he's terrible at it. Just that there are lots of others that have been better, though over a small sample size.
Total teams correctly selected:
2006-2010 (65 teams) 63, 63, 65, 64, 64
2011-2014 (68 teams) 65, 67, 68, 67
It seems impressive. Consider that by the time selection sunday comes - removing AQ's and locks from the equation - you're left with about 10 spots and about 15 teams. By default you're guaranteed to get 5 of those correct. Assuming you get two or three of the remaining 5 correct, you end up at about 65, 66.
Overall, Lunardi is slightly above average at what he does. (out of a possible 408, he is 1.39 pts higher per year than the average bracket in bracketmatrix)