What are the advantages of dispersing; a paper by Kuno explained and extended

Contrary to Kuno's (1981) contention, dispersing does not help and individual to get a larger average progeny in an unpredictable and heterogeneous but nonlimiting environment: average progeny is exactly equal for (partially) dispersing and nondispersing populations. However, the geometric time averages of pro-capita reproduction as well as geometric averages over replicates of final progeny size after a fixed number of years differ, just as Kuno asserts. Moreover, if populations of the two types are grown in mixed culture it is the disperser who will win in the long run. This even applies if dispersal means the incurring of some additional mortality. Models with partial dispersal are much more complicated to deal with than models with either a complete redistribution each generation or no dispersal at all, contrary to the assertion of e.g. Venable and Lawlor (1980). Partial dispersers will win from nondispersers, but the optimal amount of dispersal unfortunately seems to depend sensitively on the details of the model specification, except that it has to be small if the number of independent patches is large.