For the 10-100% routine it is suggested to ensure that this happens in one shot. I think @Coulomb gave an explanation elsewhere as to why.
Actually, I really don't know about that now, whether I said one way or the other previously. I suppose it depends on exactly how the BMS firmware is written, but now I'm thinking that most likely, interruptions to the charge should make little if any difference.
Thinking specifically about the LFP chemistry, which at present means the Standard Range models for MGs. When the BMS sees low battery voltage, it can make whatever correction to its estimates that it sees fit, and similarly at the top end. It doesn't have much to do while in the middle State of Charge regions.
For the NMC batteries, it could be adjusting its estimates all the time, but it probably restricts itself to regular adjustments, say once per day.
For both chemistries, it has to take into account temperature and charge/discharge history, so maybe it sets a flag at a certain time, and if it happens to be charging or discharging at that time, or in the last hour or two, then set another reminder for an hour or two later to check again. I wonder what happens with taxis and the like, which might rarely have several hours when not charging or discharging. Presumably it just does its job in the rare quiet periods, and that's probably enough to estimate the state of health well enough.
None of this seems to be particularly sensitive to interruptions, it seems to me now. But this is just my speculation; I'd be happy to hear from anyone who knows more.
The above was all about estimating state of health, i.e. the battery's present total capacity. Estimating state of charge (instantaneous capacity remaining right now) is probably different.
