My 1981 publication studies A064380 with the quite natural convention A064380(1)=1. So a(1) could alternatively be defined as 1. By the definitions, it is clear that A064380(m) >= A000010(m).
Theorem. For every n >= 0, the equation A064380(m) - A000010(m) = n has infinitely many solutions.
REFERENCES
V. S. Abramovich (Shevelev), On an analog of the Euler function, Proceeding of the North-Caucasus Center of the Academy of Sciences of the USSR (Rostov na Donu), 2 (1981), 13-17.
Vladimir Shevelev, Multiplicative functions in the Fermi-Dirac arithmetic, Izvestia Vuzov of the North-Caucasus region, Nature sciences 4 (1996), 28-43.
