I guess the two small ferrite U pieces form a toroidal loop, which would give pretty good coupling between the coils. I would prefer less coupling in a line filter, to block both differential and common-mode spikes.
Not much ferrite there, and lots of wire. Probably saturates easily with DC load current.
Regarding saturation and stored energy, when I was thermal testing the Schaffner I connected it to a bench supply, 2 volts and current limiting at 3 amps. When I disconnected, I got a nice little spark, from stored energy in the inductor.
I've never calibrated sparks. How many nanojoules does it take to make a visible spark? I used to make CD ignitions for motorcycles, and I recall that 50 mJ was a fat, noisy spark.