Understandably, the relative atomic masses of isotopes are often not close to whole numbers as they're adjusted for isotope abudandancy eg. $A_r(Cl)\approx35.45$. However, wouldn't one expect the relative atomic masses of individual isotopes to be rather close to whole numbers, given that they're merely the sum of the relative atomic masses of the protons, neutrons, and electrons within the atom and that the masses of protons and neutrons $\approx1$?
I.E, we have $Rb-85$, which has a literature $u=84.91$. Attempting to calculate $u$ given proton, neutron and electron count: $$ A_r(p^+) = 37\times 1.0073\approx 37.2701\space Da $$ $$ A_r(n^0)=48\times 1.0087\approx 48.4176\space Da $$ $$ A_r(e^-)=37\times 5.4858\times 10^{-4}\approx 0.02030\space Da$$ $$37.2701+48.4176+0.02030=85.708$$ The answer I get is $85.708$, whereas something close to $84.91$ is expected.