The Urban dictionary defines plausible deniability as:
A condition in which a subject can safely and believably deny knowledge of any particular truth that may exist because the subject is deliberately made unaware of said truth so as to benefit or shield the subject from any responsibility associated through the knowledge of such truth.
In cryptography deniability refers to inability to prove that the encrypted or plain-text message exists at all.
One important thing about safety of an encryption tool is ability to resist a direct attack when an adversary has captured a device often together with its owner. It is well-known that the weakest part of a cryptosystem is a human. To break an encryption it is much easier for an adversary to “break” a human part of the chain, so it would be much safer not to reveal the encryption, pretend that it does not exist altogether.
You may try to deny that you have a key, a password or whatever else in case you got captured by a legal authority. However in some jurisdictions such as UK, not giving up the keys is a crime. You'd be better off giving up the keys and deny that any other keys exist (however, that might not help and you will stay locked). If you hold the other keys and wouldn't reveal them, you are in a trouble – there are plenty of ways to prove that the hidden files exist. The operating system, third-party software, such as text editors, search assistances leak information about a location of files, time stamps, disk serial numbers or even the plain-text contents of a file that may prove the existence of the encrypted files. Lying to a law-enforcement officer, for example in the USA, is a crime.
But imagine that you've got captured by an adversary who doesn't care neither about legal consequences nor about you personally. There is such a thing as a rubber-hose cryptanalysis. Sooner or later you will have to reveal the information. But it depends what information you have to reveal. There are no proofs of how many more keys are out there, so there always is a possibility that you hold one more key. Everybody knows that and it may lead to continuous torture.
All in all, a deniability might work in certain situations, but in a really serious case it makes things much worse. You might be tortured even if you don't have a key. There are no reliable ways of telling that there are no more keys. Or, on the legal side, there are a few options to keep you locked, starting from “spoliation of evidence” and ending with “terrorist suspect”.
Multiple login accounts. Rubber-hose until done. That may work out if a rubber-hose approach is inappropriate for any reasons. But, there is one important point. Login password won't reveal any really important data. Well, the adversary will know your e-mail account, address book, notes and gallery. But the contents of the messages will remain safe unless you saved them locally to secure notes.
Key expiration. This is the best solution since you don't need to lie or hide anything, you just tell the truth that there were the keys but not any more, they are vanished. Moreover, there are no ways you can recall the keys because of their complexity and absence of any human-readable forms of those keys. They can try to restore the erased and overwritten keys (but keep in mind, that no matter how many times the information is overwritten, due to the nature of flash memory used in a device, a low-level research might reveal the deleted information).
SenseMail utilises both of those approaches (Well, expiring keys will be released in the next version of the app). It's up to you which one to use... That's also the reason why SenseMail doesn't use any biometric stuff.