Charge is defined in units of Coulombs, which in the Reciprocal System have units of "s" (1 ampere x 1 second = s/t x t = s). Omitting the fudge-factor constant, the equation is unitless, since the distance is also has units of space. They introduce "k" as a magic constant containing units to make the equation work.
In researching Basic Properties of Matter, Larson indicated that the cgs-based statCoulomb would probably be a better choice of units for measuring the electric charge, rather than the SI-based Coulomb. In conventional science, the two are considered to have the SAME units, with just a constant supplying the difference in magnitude:
1 C ↔ 2997924580 statC ≈ 3.00×109 statC (Wikipedia on StatCoulomb)
BUT, then it goes to define the units for a statCoulomb like this:
1 statC = 1 g1/2 cm3/2 s−1 = 1 erg1/2 cm1/2.
Rather bizarre combination, so let's break it down into natural units of space and time:
It has units of the square root of time--not even CLOSE to the Coulomb units of space! But, if you use statCoulombs in the equation for the force of attraction/repulsion between charges, it works:
When correcting the units for statCoulombs and using them as the magnitude of charge, the equation balances nicely with units of force. This would indicate that the Coulomb is a quantity of electrons, not the electric charge. Larson had come to a similar conclusion regarding electric charge q (versus Q), making one have units of space and the other units of energy (t/s) to resolve problems with the Farad. Appears he had the right idea, but just implemented in the wrong place.