>But it doesn't make a lot of sense to me since influenza will be mutating to escape the immune response
Not all mutations are equal; the flu has antigenic factors that rapidly change in response to evolutionary pressure that don't significantly affect the fitness of the virus. If you're able to target another element of the virus that is conserved, or target a large segment of the possible antigenic conformations in a single vaccine, you'd force immune evasion mutations to significantly lower the virus's fitness, which is a very big deal.
The 18 variations are like a deck of 18 cards, the virus 'shuffles' one into use, next year another works better. The multivalent vaccine is for all 18 cards, evolutionary escape is far less likely, if not impossible.
Not all mutations are equal; the flu has antigenic factors that rapidly change in response to evolutionary pressure that don't significantly affect the fitness of the virus. If you're able to target another element of the virus that is conserved, or target a large segment of the possible antigenic conformations in a single vaccine, you'd force immune evasion mutations to significantly lower the virus's fitness, which is a very big deal.