I was sick all last week and completely missed this.
Anyway, the previous April record for a PlayStation system in the U.S. was the 411k sold by the PS4 in 2020, which beat the previous record of 332k sold by the PS2 in April 2005. Incidentally, that was the PS4's second-best non-holiday month ever (Sept. 2014, Destiny's release month, takes the #1 spot). It was up 177.7% year-over-year, and was also up 191.5% over Feb. 2020 despite February being on average the strongest non-holiday month of the year and April being one of the two weakest months on average (along with May). It will forever be a mystery how it managed to accomplish this, especially considering how the XBO also managed inexplicably massive growth. Curious.
Moving along, assuming that 98% of hardware revenue went to the main three consoles, then the PS5 being #1 in revenue means it had to have grossed no less than $120M. Assuming an ARP of $520, that would yield a minimum sales figure of at least 231k, definitely well short of the April record for a PlayStation system. Of course, that's just the lower limit. The actual number could still be a good bit higher. Figuring out upper & lower limits for the other systems is a lot more difficult, though. Without solid numbers, we have some pretty big margins of error to work with. The revenue split could have been very close, say, 35%/33%/32%, or it could have been a 45/40/15 split, or a 40/35/25 split. At the very least, we do know that since the PS5 was #1 in revenue and the Switch #1 in units, assuming an ARP of $525 and $300, respectively, that does put a cap on the Switch-to-PS5 unit ratio, which would be just short of 1.75-to-1. Of course, since we don't know the exact SKU split, even that has a margin of error attached to it, but it should be a decent approximation.
Hopefully we get some more leaks eventually.