Uh, where are you getting these numbers?

Since when is a 300MHz NMR capable of resolving peak ratios down to 5 decimal places? You know the resolution on the machine is much less than that right? Any impurities past <1%, especially on a 300MHz, are going to be in the baseline. You can't read more than 99% or >99% purity with an NMR, at least no one will take you seriously if attempt. You need to use LC-MS for that.

And on that note, even your integrations are only to 2 decimal places? so how is it even possible you getting digits past 1%? (0.01)

Any % purity or product ratios determined by NMR is ONLY reported to a single digit. 99% is as much as you can say or a clean spectra with no visible impurities you could report >99%.