@Lee_T. In the old all drum systems, the master cylinder was sealed. In disk brake systems, as the pads wear down, fluid is retained in the calipers to compensate for the wear. That lowers the fluid level in the reservoir. The reservoir is not pressurized when the brakes are applied, only the rest of the brake system.
There is a hole in the reservoir cap that allows air to enter the reservoir so it doesn’t create a vacuum.
You took it out of context. He’s saying that the one end is pressurized so air is not coming into the reservior from that end and goes on to say the reservior is also sealed by the expandable diaphram on the opposite end.
Air does not enter the reservoir in any significant amount as that is the purpose of the sealing diaphragm. It gets sucked downward as fluid moves toward the wheel cylinders/calipers over time due to pad wear.
It’s true no seal is perfect and I’ve had systems that were more “lossy” than others. But by and large, not much air intrusion is going on if, when you break it open to service the system, those diaphragms are extended into the reservoir- meaning they have been doing their job to displace the lost (re-distributed) fluid and prevent air from being sucked in.
I never top off my brake fluid in any vehicle. It just opens the system to air (albeit for a brief period) but more importantly, I lose the ability to gauge pad wear in the most simple way possible- by examining the fluid level over time.
@TwinTurbo, if you examine the cap on your master cylinder reservoir, you will find a hole about 1/16" to let air in.
Thank you for that TwinTurbo. I thought I had communicated that but the next thing I know shadowfax is getting snarky with me and I don’t even know what he’s arguing about. Then keith comments about the vent hole in the cap without getting the part about the diaphragm.
Maybe in the old days master cylinder caps didn’t have diaphragms and were vented to the atmosphere and that’s why brake fluid deteriorated so quickly. Maybe corrosion inhibitors have improved through the years.
That’s to let air in above the sealing diaphragm so the diaphragm will go down. But air does get in past the diaphragm. Again, spacecraft seals are a lot better than anything you’ll find on a car, and air gets past them too. It’s not much air, but over sufficient time it’s enough that it’s a good idea to replace the brake fluid periodically.
Air can also diffuse through the rubber (flex) hoses and water can get past the sweeping seals of the calipers. It’s minuscule but can add up over time. I think the bigger issue is that the corrosion inhibitors in the fluid break down over time and it can become corrosive. That’s why the better test is to check for copper suspended in the fluid as copper is the first to get attacked.
All that being said, I’ve never been one to proactively replace on the manufacturer’s schedule. And I never have failing calipers etc as a result. Must be clean living 
My sixties era GM musclecars all had rubber diaphragms in the MC. After many years, the rims of the cover and MC get corroded and then they start being more problematic. But at least initially, they were sealed by design.
There is no sealing diaphragm any more. Haven’t seen one in MANY years.
Some vehicles have a diaphragm in the brake fluid reservoir cap, some Chrysler products still use the same cap that was used in the 1990’s.
I am not familiar with Subaru but most Toyota vehicles that I work on have vented brake fluid reservoir caps.
I just took the cap off the MC of my 14 Legacy and it looks exactly like the cap you show here, even the color is the same. Since they started making see through reservoirs, I hardly ever remove the cap unless I’m flushing the brakes.
Now I’ll bet your thinking that I am backing off my statement, I’m not. I still have not seen a sealed diaphragm in a cap for a long time because dead center in that rubber diaphragm is a tine slit that lets air through. You have to press on the rubber to see the slit. It is almost invisible if you don’t.
Now I have seen clutch master cylinders that have a sealed diaphragm, my 02 Saturn had one.
I’d be willing to bet that slit is in there as a safety vent in the event someone never checks their brake fluid, it gets too low, the diaphragm is fully extended and to prevent air lock, it will breach the slit as a last ditch effort to stop from starving the system of fluid. The resistance to the diaphragm extending is less than that required to get air past that slit…
In addition to the excellent illustrated answer offered by Shadow (I was simply going to ask you where you thought the fluid came from to keep the lines full and where you thought the fluid went when the level dropped), remember that the elastomeric components entrapped in the brake system (see my earlier post) break down over time and contaminate the fluid. That alone is reason enough to occasionally flush the system with fresh fluid.
If you choose to not ever flush your system, I support your right to do so. But recognize that your motivations are not that the fluid stays fresh because the system is “sealed”. That part goes against the system design, which allows the fluid to be drawn into the brake system to keep the lines full… and it ignores the deterioration of the mentioned elastomerics. 
Anybody here got a photo of well used, contaminated brake fluid handy?
Heck yeah! 150k miles. But how do I attach? Lol
google to the help ? 
switch to “Images” tab, could not do it with LMGTFY
I’m doing this on an iPhone. Do not see an images tab.
Also, my wife drove the 150k Buick unit tonight so I could change the oil in her 95k mile Highlander. So it would be 95 k mile brake fluid.
My real opinion - brake fluid attracts moisture. But if you’re changing it every 30k, that’s probably overkill. I’d change at 100k, but I’m lazy. I don’t give much of a chit about the Buick, and I’m willing to bet the toyo will be alright till 100k.
Simply tap the little line in your reply box with the upward aiming arrow and a window will open allowing you to select the image. Then click on the image.
My drum brakes on all 4 wheels truck’s MC lid has a very tiny hole in it, presumably to allow air in above the flexible diaphragm. Not sure if shoe wear is involved but you don’t want the air in the MC to pressurize due to heat in the engine compartment. I didn’t even realize that hole was there until one time I turned the lid upside down when cleaning it using some fresh brake fluid and a brush, and the fluid came dribbling out.
Re whether low fluid level in the MC can result in air getting into the brake lines … well, there’s definitely some air above the brake fluid, so it seems to me that could happen, right? If the brake fluid level got so far down that air filled the cylinders the pistons would push that air into the brake lines as the brakes are applied. I’m not sure whether that air would stay in the lines tho, or just bleed out through the MC later.
Man, I don’t see a little line in my reply box. Took a pic, have it ready. Again, this is 95ish k mile 2013 Highlander, if we get it working.
And I’m on an iPhone.
8 yr older Buick looks worse, btw.
And I’m not advocating never changing, again. I’m no expert.
That would be my guess as well, slit in the diaphragm as a failsafe.
Not sure if the image works. This is the master cylinder of a 2013 with 95k mikes, brake fluid never changed. The Buick looks worse, but it’s cold out lol. I need to change it, has some sea monkeys floating in it.
I suspect age rather than mileage is the determining factor in change interval, which would make sense I guess.
