Yes, I hear you, that's why I used real 2.7V LDO regulators, not forward biased diode. As of EV error, I must say two well calibrated meters may easily differ by this amount - highly dependent on how you point the meter to the subject, spectra of light (if sight sensors are made with different technologies - it's impossible to compensate exactly, to make one meter function as the other) , and few other variables. In short - if you have more than one meter, you likely have up to 1/2EV readout difference between them already, whether you want it or not.