In the first example the battery is being very heavily discharged (high discharge current) to a very deep depth of discharge. This severely shortens it's life. On the face of it, it would seem that doubling the size of the battery bank would double the life of the bank as a whole. However this is not the case. Doubling the size of the battery bank could increase the life of the battery bank as a whole by say 3 times. So double the initial outlay (twice as many batteries to buy) results in a saving of 50% (they last 3 times longer).
In the second example, this batterybank is going to last a long time. It is being well treated, lightly discharged to a reasonable depth of discharge and charged at a sensible rate. However, reducing this battery bank by 50% (to a single 100Ahr battery) would perhaps only reduce the life of the bank as a whole by say 30% because the single battery would still be being relatively well treated. So in this case a saving could have been made by buying half as many batteries (so half the initial outlay) and getting a battery life of 70%. A saving of 30% in monetary terms.
This explains it in plain English. It can be shown mathematically, by graphing the cost of each used amp hour against the initial monetary outlay. The result is a graph with a peak in the middle at, you guessed it, 50% depth of discharge.
Discharging deep cycle batteries to 50% results in the most economical use of the batteries in terms of battery life and monetary outlay.