I would suggest that you take the approach of isolating variables and testing them. Chocolate is a very mysterious and challenging medium. First I would temper some chocolate and mold it the same as you always do. If you are certain that you have good temper then I would test the rate of cooling (cooling times and temps). Place the tempered molded chocolate on your counter (room temp.), some in the refrigerator, and some in the freezer. I know you can get some condensation when re warming the chocolate out of the freezer but you said the humidity is low where you live so shouldn't be that big of a problem. Test multiple samples in each location changing the exposure time for each. Record this information and keep the bars separated. Observe the samples over a couple of days or longer and pick which one best represents what you are looking for. This will hopefully allow you to isolate the actual problem and make the correct adjustments.
I know it was mentioned in earlier responses about shocking the chocolate and causing bloom by setting the chocolate too fast. I would tend to lean the other direction with my experience and say that it is not cooling fast enough and actually un tempering the surface of the chocolate with the latent heat remaining in the chocolate. With hyper cooling you usually get some visual identifiers on the surface of the bars such as cracking or fracture lines. Just my two cents. Happy trouble shooting;)