It quickly becomes a question of economics.
Let?s do some quick assumptions ?
? The average level of interesting spells (healing, speed growth for large areas?) is 10.
? 1 in 100 is a spell user
? 1 in 10 of spell users are capable of casting level 10 spells
? The average number of PPs available for such a spell user is 30. (1,5, average level 10, x2 multiplier)
? 1 in 3 spell users have spell lists relevant to the population.
Doing the math, we have 1 person of 3000 (1/100 x 1/10 x 1/3) capable of casting spells interesting for the general population. Those people can cast 3 relevant spells per day. (30 PP, level 10). In a perfectly fair world, this would give an allocation of one spell per 1000 people per day. And I truly believe that the above numbers are too high, but let?s run with them.
1 person in 1000. So, assume you are a priestess of Eissa. You can cure a homeless child?s pneumonia and then let it out on the street to catch another bout in a few weeks, or you can cure a nobleman?s sprained ankle and charge him enough money to run an orphanage for a year. You would probably hate the mathematics of it, but in the quest for doing the most good you would heal the nobleman.
Even in this very optimistic calculation, I don?t see the general population seeing much magic outside of festivals and such.
Note: I think that 1/100 spell users is too high for most worlds, 1/10 at level 10+ is also probably too high. 1 in 3 having spell lists that are interesting also high.