Well, if you could get a bottle of bacteria to have an oxygen source for sufficiently long enough and get it from the factory to the customer's tank within a month they could work without refrigeration. Without food, you loose arround 2-12% (depending on the author of the scientific paper you're reading, generaly in the 2-5% mark though) per day. This is a slow die-off, so you can get *some* live bacteria in the bottle provided you have oxygen in the bottle at all times...
Without an oxygen source or refrigeration, you will loose about 90% per hour.... This means from factory to tank in less than a day to work. Not gonna happen, though this is the scenario we see with most "bacteria in a bottle" products...
These products are good little money spinners for the manufactueres, and not too bad for the LFS. They don't state how long the bacteria will last for, or the species contained. This makes the product very hard to prevent from being sold, even if incorrect, as it hasn't been miss-sold if they haven't said that it will all be dead on purchase, or that it's the wrong type... They can instantly cycle a tank, but you have to get from factory to customer's tank within the hour for them to work, and they have to be amongst the few manufacturers using the correct filter bacteria...
All the best
Rabbut