A fellow I know used to have a private plane that was pressurized. He had an intermittent leak problem somewhere and had a heck of a time getting it resolved. I do believe you have to avoid leaks to a fairly significant degree...
Finding and resolving individual leaks can be time-consuming and annoying. That being said, leaks are a very small and infrequent issue with aircraft pressurization systems: it's neither difficult nor usually threatening. I'm involved with the operation and management of two pressurized general-aviation airplanes, and have been for years... it's just not a problem. Takes work and time, but does not introduce any significant risk.
While the risks of decompression are very real, providing a safe environment within the Hyperloop shouldn't be too hard. The pod's passenger and baggage compartments are a sealed bottle, essentially: you load the people, you close it up, you're done. If you have a minor leak, you use some of the reserve air to compensate and maintain pressure. All you have to do is define the loss of air (easily measured in air mass consumed from the reserve) at which a minor leak flags the pod to be taken out of service for maintenance after this particular trip. That's not hard at all, and it would cover some large (way over 99%) fraction of cases.
You'd only need the masks if the leak was significant enough that the reserve air supply couldn't keep up and the pod ended up losing 31% of its air pressure (down to 10psi/70kPa, roughly the equivalent of 10,000 feet of altitude) within the length of a single trip. That's a hell of a leak, and should hardly ever happen; especially since you can design into the pod the ability to resupply itself with large amounts of air from the reserve to reduce the odds. Even here in the "minor leak" realm, the probability that any given pod will need to drop the masks is minuscule: if it happens four times a year on the SF-LA loop, which would still be (IMHO) unrealistically high for a well-designed pod traveling inside a protected tube, you're talking about a 0.0000025% chance (1 in nearly 400,000).
In that rare case where you did have to drop the masks at 10psi, you still have a big margin between that and 2psi/14kPa where things get life-threatening. The system operator would set a hard limit at which emergency repressurization of the tube is undertaken, and I'd say (for the purposes of this discussion) it would be something like 6psi/41kPa to enable the tube to reach a reasonable pressure of at least 3psi/21kPa before the pod ever gets to a life-threatening situation. The key problem to solve here, which is not egregiously difficult, is to determine how quickly the tube can be brought up from its standard operating pressure to 3psi/21kPa without incurring unacceptable effects of shockwaves, friction heating, and so on, and then set your safety parameters that way.
Given a good design effort, I just don't see loss of pressure as being a highly-difficult problem. There are other things, like the whole expansion-joint issue, which are thornier.
- - - Updated - - -
In the very unlikely scenario where a part/piece of pod is broken off and causes damage on the next pod that disables it and causes a leak... If you are trapped mid way, you are in hostile deadly environment trapped in steel, and unlike a plane that could dive down to normal pressure, though normal environment is a few feet away you are trapped, I dont know if the pod door even open, and even if you could exit the tube environment is deadly, and its not like there'd be an exit 20 feet away, it might be miles away.
If something happens and a pod is immobilized many miles down the tube, what is the emergency response procedure?
Remember the pods are also equipped with wheels to drive themselves forward, and the tube can be repressurized so it's no longer a lethal environment, just one where the pods can't go full speed. Foreign-object damage such as you mentioned is a major concern during design, to make such a "piece breaks off a pod" scenario nearly impossible. In the strange case that it does happen and does leave a pod stuck, Steve's right and the nearest escape hatch is probably a few pylons away.