Max, a waterproof breathable membrane like goretex will allow gases to pass through it, but won't let you blow through it. The ability for fluid to flow does not mean that it will let it flow at a high rate. Blowing through a filter is "high rate" and seeping water is slow. You confused rates with absolutes. Look at all the technical reports and threads on this site relating to breathability and air permeability, similar concepts hold true for water filters.
If you want to explore the fluid dynamics it has a little to do with laminar flow and boundary layers. Fluids are still immediately next to a surface and flow faster the farther from the surface you move...so tiny tubes/holes reduce the area that air can flow quickly compared to larger diameters. Hence more efficient to have big water pipes than a bunch of tiny ones. Also viscosity of fluids has a damping effect related to the velocity of movement. The harder you blow the more resistance you'll have. Hence why you can't blow through a good filter but water will seep through it. I'm sure sealing dynamics play a role, IIRC the "effectiveness" of a seal between two surfaces is proportional to the square of the distance between the surfaces, ie diameter of tubes, and linear to the length of the tubes. That's why it's harder to blow through a long pipe than a short one.
Just because there are millions of tiny holes does not mean there are enough gaps to allow a "breath test" to work regardless of integrity. Now if we know that a sawyer squeeze does not allow air flow when new, and it starts allowing air flow after damage (freeze, drop, etc) we can say that a gap large enough to allow breath through has been created. This gap is most likely far larger than the items we are trying to filter so replacement in necessary.
This test is highly accurate, unfortunately not very precise. There could be micro damage to the tubes that don't allow breath through, but do allow water borne pathogens.
The MSR and similar filters that are "hollow tube" in design basically work by having thousands of tiny ceramic (ie rigid) straws glued together. When they freeze the ice expands and due to the inelasticity of the ceramic, it can fracture. You have a hard enough freeze and it can shatter enough of the tubes to ruin the filtering capabilities. The Sawyer Squeeze uses a different technology. First they use flexible tubes that can stretch. This should mitigate much of the "cracking" traditional filters suffer. Second they do not filter through the tube directly. Each tube is a permeable membrane with micro pores smaller than the specified micro filter sizing. Think of making loops with straws, then poking pinholes all along the length of the straw. You fill with water from both ends of the straw and the water seeps out the pinholes on the sides.
Unfortunately we don't know, and Sawyer hasn't studied, how their material behaves when frozen very well. Just because it is flexible doesn't mean the ice can't rupture or stretch out the micropores. Also depending on where the ice forms it could cause mechanical failures in the attachment of the tubes, similar to severing a tube that could allow dirty water to flow through easily.
It's just a matter of your risk tolerance. I slight freeze and you may be fine or willing to risk the filter on cleaner water source but a hard freeze could likely be problematic.
If you REALLY want to know, freeze your filter, study it under a microscope and then, sterilize a jar, filter some cow dung riddled water and culture the "cleaned" water on petri dishes. Repeat for about 30 filters (more if you're worried about contamination errors or are OCD) and you'll have a better idea of whether the filters survive. Or just keep it in your sleeping bag at night ;)