If the period for which the message can't be read it is < 2 weeks, then yes. This forum is littered with proof of work time based cryptography so I won't rehash them.
In reviewing the questions /answers here, no one seems to discuss that for short periods (<2 weeks say) it is entirely possible. The 2 week limit is a soft limit, but that's the difficulty index recalculation period for the bitcoin currency. bitcoin mining works and is proven to be an un ircumventable proof of work procedure. It must also be one of the most studied proofs of work in the world. It seems to make intuitive sense, that you might set a hardness parameter (much like in scrypt) based on current processing technology that will take up to 2 weeks to crack. You might set it for only 2 days, but there will be many applications where that would be adequate. Needless to say that the proof of work algorithm should not be able to be paralleled, perhaps something akin to the Rivest puzzle. Today you'd probably target the Intel Xeon E7 chip.
Just as bitcoin does, the 2 week limit eliminates the risk of mathematical /technological leaps in hash computation. We can debate the exact delay period, but that will have to include a trade off between risk of early decryption vs. potential damage. At a long stretch, this might be comparable to the standard definition of a hash security parameter, but measured in seconds, rather than bits.