Biologically speaking the model answers the question 'Can the phenomenon of flocking birds ( or other animals ) be explained by decentralised activity?' ( i.e. not by an overall controller co-ordinating the behaviour of the flock). The answer based on the boids simulations is a quite emphatic, 'yes it could be' ( although note these simulations alone do not prove that flocks operate without centralised control ).
Further, boid simlulations provide a good and clear example of what emergence is in nature. The boid simulation has the additional advantage of producing behaviour in terms of emergence that looks rather more life-like than in the Game of Life ( although both simulations obey only a handful of rules).
However, artificial life-forms, based on boids, can also be used to examine problems that hvae nothing to do with biological phenomenon...
In particular computer simulations of this kind have been used to examine the bizarre world of traffic flow ( if you've ever spent time on a motorway in Britain, you'll know why use the word 'bizarre'!) One such example is the simulation implemented by Herr Treiber at Stuttgart University, http://www.theo2.physik.uni-stuttgart.de/treiber/MicroApplet/.
If this sounds odd, then consider altering the boid code a little to add different factors of agression which make the boid-cars more likely to speed, swerve or drive on the hard shoulder. If successful such a model could provide important insights into good traffic management strategies...