We won't be able to know precisely how the satellite data matches up until next year at the earliest when the 1986-2015 base period becomes available. The OISSTv2 record began in November 1981, and even though you could technically run ONI data on this dataset using a 1982-2010 or 2011 base period, this will yield slightly different ONI values from ERSSTv4 & v3b that are currently calculated against a 1981-2010 base period & at that point, you would be comparing apples-oranges. There are several other relatively independent SST reconstructions & sets that are available and ONI data could be derived from each, including Kaplan v1, Kaplan Extended v2, HADISST1, HADSST3, IOCADSv2.5, COBE SSTv2 (the COBE data has been extremely hard for me to get, I have to compute ~980,000 1x1 degree boxes in the NINO 3.4 region to get the monthly temps thru GrADS). The KNMI Climate Explorer is a great place to start.