The simple fact of the matter is that we don't know how good writable Blu-Ray discs are yet -- there isn't sufficient data to say whether or not they'll stand the test of time. Accelerated aging tests can only get you so far. They certainly make a degree of practical sense in terms of cost, shipping, and so forth, but as a long-term solution, right now, you would have to be willing to engage in a rigorous program of periodic disc duplication in order to stave off any potential "bit rot". At least until the medium is proven, or its archival qualities are properly understood. (And I's stick to single-layer, since clouding -- obstructing the deeper layer -- is one of the more probable failure modes.)

Of course, the same can be said for any digital storage medium. Magnetic domains aren't forever either, so periodic rewriting is essential. Then there's always the question of long-term readability -- there was a time when ZIP disks and magneto-optical storage sounded like a good idea, but now we have the problem of sourcing readers for those disks. How long will it be before it becomes the next best thing to impossible to find an interface for an EIDE/PATA hard drive? SCSI? When will SATA be superceded? Or USB? Even file formats change over time, so there's no real guarantee than twenty years down the road your files will still be readable (this will be less of a problem for well-entrenched formats like JPEG or TIFF than with any proprietary format, but you never know).

The advantage to magnetic storage is that the storage capacity of individual drives is much larger than optical disks, so when you need to re-archive (and you will need to) there's a lot less donkey work involved in the process. I'd much rather swap 1TB hard drives than 25GB optical discs any day.