Lots of these engines suffer lots of lean pops through the intake, so it's not one of those 'if this ever happens, that will always follow' things. Could be that there's something preventing the valve(s) from closing fully (spring retainer popped out, spring broken, crud holding it open); or the machined surface that provides the seal might not be machined quite right. It's not a rubber seal that's compromised - it's a machined brass/aluminum interface.
Given that the engine they were on before was relatively tame and the jetting that these carbs came with AND that it would only run with the choke active, it seems likely that the carbs have had more, and possibly more significant, pops through the intake than they ought.
Have a look here, and you'll see what's involved in 'fully disabling' the cold start provision. You might be able to staunch the bleeding with a good cleaning, or stronger springs over the plungers, or spacers or some other thing. I've seen some weird approaches to this (grub screws in the exit, which is in the throat just downstream of the throttle plate - kept fuel out of the stream but didn't prevent leaks to the outside world), solder blobs plugging the supply holes in the bottom of the bowls, stuff like that. I'd see if I couldn't get it corrected without modification first.
Oops: the link:
https://www.carbparts.eu/books/cold_start_elimination_kit.pdf
You'll get there. Might go crazy along the way, but that's part of the fun.
Cam
Portland