Consider the current dataset value of the elementary charge:
In[1]:= Quantity["ElementaryCharge"] 10^19 // UnitConvert // First
Out[1]= 80108817/50000000
In[2]:= % // N[#, 10] &
Out[2]= 1.602176340
whereas the correct "New SI" exact value (times 10^19) is:
In[3]:= 801088317/500000000 // N[#, 10] &
Out[3]= 1.602176634
Apparently, one significant digit has been dropped, which affects all the corresponding unit conversions.
Data source: https://physics.nist.gov/cgi-bin/cuu/Value?e
