Well your not going to like this answer, but here we go.
The tolerance for a Grade 2 set of gage blocks is around 4-6 uin depending on exactly what procedure is used by your cal provider. From the looks of the data your cal provider is using 5 uin as the tolerance because 5 uin was in spec in 05, 06, and 5.1 was OOT in 09. The tolerance for the 2 different blocks is the same regardless of procedure.
Here is the fun part. The BMC for gage block comparisons for most accredited labs is about 3-8 uin. It depends on their standards, and comparison meathod. This means that either of those two blocks could acutally be in or out of tolerance. Due to the Measurement Uncertainty being as large as the deviation from nominal, it is immpossible to say for sure if the gage blocks are in or out of tolerance.
Different labs handle these problems many different ways. Some will call the UUT in or out of spec without considering the MUC. Some will not call the UUT in or out, and just tell you the measurement. Some will fail all UUT's that miss the spec - the MUC. All of these, and coutnless others are acceptable ways of handling complaince statements. There are requirements of the lab in how they handle compliance statements, but these vary slightly form one accrediting body to the next. There most likely is a note somewhere on the cert, of the qoute that states the labs MUC. Some labs also have an explination of the compliance statement process on the cert.
In the end the lab that made the measurements probably made good measurements. There is a chance that either of those blocks do, or don't meet spec's of a grade 2 set. There is a chance that alot of the other blocks may or may not meet grade 2 specs either. What matters is if any one of those numbers plus the MUC of your cal provider could have a negitive effect on one of your measurements. If you only use these to calibrate hand tools 7 uin is not going to effect anything anyways. if you do gage block comparisons of your own then you may have a problem.
I have to add this part.
Your cal house most likely did a fine job here. All calibration labs will run into simular issues on grade 0,1,2 block sets. There is no way to maintain a 4 to 1 TUR on some types of equipment anymore. This is exactly why accreditied labs are important, and they report MUC to their clients. That way you do have all the numbers needed to see how the measurements effect your process.