Evidence Based Beekeeping

I was recently asked, by a friend on Twitter to give my opinion on this article from the University of Sussex Apiculture Laboratory. I am just an amateur beekeeper so certainly would not claim to have anything like the same knowledge and facilities as those at the university so this is purely my own personal opinion based on 8 or 9 years of beekeeping.
I whole-heartedly agree that beekeeping does need scientific research and welcome any initiatives which will cast new light on some of beekeeping's mysteries. However I would suggest that it is far from correct that much of what we do as beekeepers is based on nothing more than traditional lore and custom. A great deal of research has been carried out using scientific methods over the many years that people have been writing books about beekeeping. It is just that the equipment and monitoring facillities available to the professional research scientist are far more powerful and sophisticated today than anything available at the time that Langstroth was busily revolutionising the craft.
I believe that most beekeepers would aim to carry out evidence based beekeeping if the evidence was available and was interpreted in a way that was applicable to their particular circumstances. However, the variables involved, have highly complex interrelationships.
For example, the survival of a queen being introduced into a hive will be dependent on several factors:

1. Strain of bee that produced the queen
2. Genetic heredity of queen
3. Strain of bee in target colony
4. Time of year
5. Status of target colony - how long queenless/existence of brood/availability of stores/proportion of nurse/housekeeping/foraging bees
6. "Mood" of colony - have they been recently disturbed/will they be disturbed/more or less agressive
7. Local climatic conditions
8. Local availability of forage - is there a flow on
9. The "Quality" of the queen - is she laying well/physically well developed/nourished

and several more that we are probably not even aware of yet.
These variables will all affect the level of pheremome responses to the various stimuli involved. The use of a cage is intended to give the resident bees a chance to get used to the chemical makeup of any queen that is introduced so that comunication can be set up before the bees come into direct contact with the queen. Without this separation it is likely that alarm messages will be propagated and the new queen will be killed. The attack message is communicated by scent so heavy smoking of the hive will also work by masking the attack scent. So the bees will not attack the new queen. This could well be effective every time because the mechanism is known and the variables in the laboratory apiary are probably pretty constant. The queens are probably raised in a standard way rather than being sourced from whatever the beekeeping suppliers happen to be breeding or importing at the moment. The arrangement of the apiary, the relative position of the hives, the location of external stimuli will be fairly consitent. The important thing here is that the research predicts the probablity of a particular outcome given a particular method. The value of the research should be in the identification of the variables and an assesment of there interractions and relative validity rather than determination of a particular method on the basis of probability.
The evidence suggested by the results of any particular scientific experiment will need to be replicable by beekeepers in practice. What I am trying to get accross here is that before beekeepers can practically follow an "Evidence Based" approach to beekeeping they will have to understand which variables are significant and be able to replicate conditions that will allow the recommendations to be followed.
Scientific research and the dissemination of its findings costs a lot of money. Recently the only big money in entemological research has been provided by the agrichemical companies who obviously will have a diferent focus. It is to be hoped that with the promised Government funding the money will finally be available to determine the relative value of all the variables involved in assessing and working with bee physiology/sociology climate/environment/chemicals etc. and inform the beekeeper accordingly.
Will I be just using smoke to introduce queens in future?
Well... if I buy a queen from most suppliers she is going to cost in the region of £40, in the height of the season it is quite likely that there will be no queens to purchase and if you can it could be up to a week before you can try again, there will have to be some pretty compelling evidence to persuade me to risk it! I will next year try to start raising my own queens so if I have plenty I may well experiment, does that qualify as "Evidence Based Beekeeping" ?