Ok... lets just look at Brute Strike, since it does nothing but damage (if there are other effects, having these be Reliable might be more important than sheer damage). And we'll use a maul wielded by a strength 18 fighter with weapon focus- hammers.
Suppose we're designing this power, and we're deciding whether to make it Reliable, or Miss: Half. The power otherwise remains as is. Also assume a typical fighter hits 50% of the time when using this power, because its usually used against higher level foes.
The expected damage on a hit is 6d6+5, which equals 26.
If the power is Miss: Half, the overall expected damage is 19.5, because there's a 50% chance of hitting and getting an average of 26, and a 50% chance of missing and getting an average of 13. Next round, you won't have Brute Strike anymore, so you'll use Reaping Strike for an average damage of 2d6+5=13, with a 50% chance of hitting, and 4 damage on a miss. I'll spare you the math, but your total expected damage over two rounds comes out to 28. If we took it to three rounds, it would move to 36.5.
If the power is Reliable, the expected damage is just 13. But, on a miss, you can use it again the following round instead of whatever else you were planning to use (still using Reaping Strike here). Your expected damage over two rounds will equal, in total, 23.825. If we took it to three rounds, we'd get 32.45.
If we were to continue forecasting and check three round damage and four round damage, the more rounds we considered the closer they would become. They wouldn't necessarily ever meet, however.
I'm not sure what conclusions you should draw from this, since the better Reliable powers have secondary effects.
