>> What I would like to know from them is how much can we really expect
>> from any kind of bee breeding program?

It depends upon the trait that you want to breed for.
Many traits involve a suite of genes.  The fewer the number of genes needed, 
the easier.
Measure that trait in all your colonies to see if there is variation, e.g, 
test for mite level Aug 15.
Breed from the "best" (this is where you can go wrong, since there may be 
unintended consequences) to see if the trait is heritable.
If the trait involves one or a few genes, there is variation in the 
population, and the trait is heritable, you have a good chance of success.

Dave's and Pete's points are well taken.
When I brought in various purported mite-tolerant stocks for a few years, I 
had no idea how my bees would perform year to year, since there was no 
consistency.  It was like the volunteer squashes that come up in the compost 
pile--you have no idea what they are going to look like!  I found it 
difficult to manage my bees, since I no longer could predict when they would 
build up, how much they would swarm, how productive they would be, how hot 
or gentle they would be, or how they would winter.
Now I'm in my second year of whittling out the ones I don't want, and 
breeding (both queens and drones) from the ones I do want.  I go through 1 - 
2 generations a year.  My potential breeders for next season are looking 
very good so far.  But for all I know, they could crash from CCD, or 
something else unforseen that I haven't selected for.

Randy Oliver

******************************************************
* Full guidelines for BEE-L posting are at:          *
* http://www.honeybeeworld.com/bee-l/guidelines.htm  *
******************************************************