The following is a paper by H. Aspden published in Lettere al Nuovo Cimento, v. 38, 423-426 (1983).
Abstract: Recent precision measurements of the proton-electron mass ratio confirm a theoretical evaluation dating from 1975 to one part in 107. The theory suggests that this discrepancy, which measurements show to be 0.833 +/- 0.436 parts in 107, may evidence a very small inequality between the magnitudes of the proton charge and the electron charge, theoretically predicted to be 1.07 parts in 107.
Commentary: The reader should here understand that the author was conscious of the fact that his theory assigned a finite spherical form to the electron, which needed to have a very much larger radius than the proton. If, therefore, the aether is a kind of charge plenum or continuum neutralized by unitary lattice charges e, all having the same polarity in our local space domain, there has to be a minute discrepancy of effective charge owing to the continuum displaced by the presence of the electron. This was not assumption, inasmuch as the very explanation for gravitation advanced by the author in the 1960s relied upon such a charge discrepancy, as seated in the microcosm of space occupied by the graviton. The problem, however, is that, in explaining very minor discrepancies at this part in ten million level of activity, one can be drawn into a realm of speculation which may detract from the quality of the primary work. It is hoped that the reader will preserve a measure of circumspection at this level of enquiry, until more is known experimentally that can confirm what is suggested.
Note here that the issue is whether the electron charge and the proton charge have equal magnitudes at this part in ten million level. We now know that a proton and an antiproton have equal mass (and so presumably equal charge) magnitude, with agreement to within 1 part in 108 [Physics World, 'Antihydrogen' pp. 44-48 (July 1993)]. This, however, can apply, because the proton space-occupancy volume is so small. If, however, the electron and positron charges are compared, on the theory of the part in ten million discrepancy of the subject paper, we should expect a two part in ten million charge discrepancy. In terms of an electron spin anomaly this would translate into a g-factor discrepancy between positron and electron somewhat larger than the one actually measured. However, the overriding effect of secondary wave resonance effects could well normalize the actions measured and make them equal.