I think you can ballpark it.
There are a lot of variables, including the production predictions, which aren't likely to be achieved and which have moved around a lot. However, let's start by figuring out how many more EV miles a year a M3 will get than a Volt will. I don't think a full BEV like the Model 3 or Bolt EV will drive a lot more EV miles than a Volt, especially the second generation Volt. We've seen this in the stats when comparing the Volt and the Leaf. If people drive 40 or less miles a day for 300 days a year, the difference in miles driven will have to come from the other 65 days. Some of those days a Model 3 owner won't drive at all because of range limitations. The Volt will pick up at least 40 or 50 miles on those days. That leaves the few days beyond 50 miles that people are willing to take a BEV. Given the people I know who have a Model S won't take it to Palm Springs, I doubt there are that many of these days. But there will be some, so let's assume the average Model 3 will put on 10% more EV miles a year.
This means that every M3 accounts for 110% of the EV miles that every Volt does. Taking your assumption that there are 4X more M3 than Volts on the road in 2019, the M3 will account for 4.4X more EV miles in that year. Hence, at the end of 2019, the M3 would have 3.4X more EV miles than the Volt (4.4X -X = 3.4X).
However, we we have to account for Volt EV miles before 2019. Assuming Volt sales were evenly distributed throughout the eight years before 2019 (2011-2018), then the Volt would have accounted for roughly 36/8 or 4.5X of the EV miles the Volt got in 2019 during this earlier 8 year period. (You can either sum the years -- 8/8 + 7/8 + 6/8 etc or just take 4.5 average years). Since 4.5X is a bigger number than 3.4X, the M3 wouldn't pass the Volt in EV miles in 2019. But obviously it would in 2020 since 4.4X is a larger number than 1.1X (what's left of the EV miles accrued by the Volt prior to 2019).
Of course this assumes the M3 hasn't accrued EV miles in 2018, which is obviously unrealistic. Also it assumes the Volt sales have been even over the years, which is likely close but perhaps not spot on. But under your assumptions likely in 2020. You may also want to check the math.
FWIW I don't think this is likely at all. Tesla is struggling to sell 100K cars a year worldwide. Assuming they're going to sell 400K+ M3 is a year just seems highly unrealistic.