I agree that Clerics would be more common than many other classes, but still... I think you wind up with FAR, far less people being able to cast Raise Dead than your math suggests. (I also think that the number of clerics of level 10 and 11 would be drastically lower when compared to lower levels).
Well, for the levels 9-11, I'm already at 0.35% of all clerics, so I'm working at a tiny fraction of all clerics. Definitely much, much less than the number of low-level clerics. Note that the numbers scale such that 97% of all clerics (or any other class) never make it past level 5.
I'm Galandris in thinking that 10% of the population being adventurers is off by an order of magnitude. I've always thought of adventurers being pretty special. The vast, VAST majority of people would be commoners, and nearly all the rest NPCs (with the usual, lesser stat-blocks that come with that - though possibly a few capable of the ritual. A very few.)
The more I think about it, the more I agree that it's too high a number. For scale:
A town of 500-1000 might have one to three adventuring parties. This does not seem unreasonable. Having 10 adventuring parties is way too many for such a small community.
A small city of 10,000 could have a couple dozen adventuring parties. It wouldn't have hundreds of adventuring parties.
So on that end of things, 1% definitely seems more reasonable. But that still leaves the question of how many adventurers have retired to normal jobs? What is the rate that adventurers leave that specialized population?
The metrics I'm using have half the adventuring population drop out every year — sometimes through death, sometimes through retirement. Basically, given a stable 1% of the population being adventurers, 0.5% of the population transitions from normal to adventurer each year, while 0.25% moves back from adventurer to normal. The difference either continue advancing as an adventurer, or die off.
Most of that 0.25% each year are going to survive for long periods of time, just like anyone else. If you start adventuring around the ages of 15-30, you probably drop out in your 20's or 30's. With a life expectancy of even 50 to 60, that's 30 years of accumulated dropouts that are fairly likely to still be around. At 0.25% per year, that's 7.5% of the population that's likely a retired adventurer, though I could see that dropping down to 5% just from normal death rates.
So I end up with 1% of active adventurers, and 5% of retired adventurers. There might be more than 5% retired, but the 5% is something I'm fairly confident about. This is actually a useful division, since active adventurers are unlikely to be working the resurrection job, so I can just work from the 5% number alone.
So: 5% (retired adventurers) * 10% (cleric proportion) * 0.0035% (level filter) * 50% (chance of accepting job) = 1/115,000.
A population of 115,000 would have an annual death rate of 1,978, of which 659 are assumed to be rezzable. So we don't have as much wiggle room, but it's still within the boundaries of viability. The upper limit of viability would be 1 rez cleric per population of 125,000.
Is it reasonable to think that there are at least 8 clerics capable of casting Raise Dead in a city of 1 million? Personally I don't think it feels out of line, but it's certainly something that could be argued.
There's still that pesky "soul must be free and willing to return" clause. If someone's gone on to the realms of Elysium, do they want to go back to mortality? And if they're in the Hells or the Abyss, will their tormentors let them go?
This whole gig only works as a life insurance system if it provides consistent results. If people are spending out 500gp for merely a possibility of successful resurrection, it becomes a lot less viable.
It certainly doesn't work if most of the population is adamant about getting to the afterlife. It just doesn't feel like that sort of population would work out very well as a civilization, particularly with respect to accidental deaths or deaths by injury or stupidity. Likewise for children.
As for the "possibility" of success, well, that really depends on the person being resurrected, doesn't it? If they definitely don't want to be rezzed, then they shouldn't be paying in the first place. If the success rate is low, then, yeah, the entire plan fizzles. And that would certainly explain the lack of common use of those spells, if that were the case.
But still...