Dear foodsafetyboy,
Slightly 
It doesn’t exactly relate to yr rather ambiguous post (eg what micro.analysis?) but you might be interested to know that thousands of tons of whole, slightly processed, fish is routinely frozen at sea in blocks. On land, these blocks are subsequently thawed in water and further processed into fish meat or fillet blocks which are themselves frozen again. These blocks are then sold on again to processors who cut the frozen raw blocks to produce thin slices which are then battered/breaded to yield fish fingers. The fingers are boxed and the boxes frozen again. And hence to the supermarket.
Basically, the bacterial control is done via “good” quality raw material and processing time/temperature parameters although the numbers will rise. Environmental contamination is controlled by hygiene.
Of course it is likely that beef has (much) more contamination in the starting material, eg fecal related, so the cumulative risks will then be greater as noted by GMO and practically well demonstrated by various historical and fairly current incidents / disasters.
I hope you don’t eat yr beefburgers raw or slightly cooked. 
Rgds / Charles.C
I think that the key factor for this example is that it is the process. The entire process is based on this practice happening.
With the original question, what is the "normal" process.
Chances are it should be "thaw what you want. Use it. Toss what you don't use".
If you are considering "Thaw it. Work out what you want and refreeze the rest. Use it." then here are the immediate issues I see:
1. how do you guarantee a fast turnaround?
2. how do you guarantee a limited number of refreezes?
3. how do you convince the sceptics?
In the eTemperature software we threw in the calculation of ice-days to predict the accumulative effects but would you be prepared to put a temperature logger with each pack that you thawed and refroze?
Cheers,
Shane
