I've come to realize that trying to model future income for scrypt miners is not easy. Let's start by using Litecoin as a base--it is really not much use to attempt to consider all of the alts out there (rather, you can use 110% of LTC profitability or something if you are on a good multipool).
Take a look at the difficulty chart for litecoin over the last month)
Prior to that, it was fairly exponential:
For now, at least, it is flat to declining (probably the last of the GPUs are being taken offline, and not being replaced by ASICs as profitability is in question and hence demand is not quite there).
Next, we need to look at the LTC price (we can consider LTC/USD prices). In 2014, LTC has trended steadily down from .03 to about .01 BTC currently. Expressed in USD, it has trended down from about $25 to about $5. This is a 75% loss for the year. Over that same time, difficulty went up about 12 times. But that trend appears to have flattened out.
Profitability ($/MH) at the beginning of the year for litecoin went from $10/MH/day (ah, the good ol' days) to a current $0.187/MH/day. That looks gloomy, as mining is only 1.87% as profitability as in January.
However, current scrypt miners hash at about 25W/Mh (e.g. Zeusminer Lightning X6, 40MH@1000W), which at $0.12/Kwh means about $0.07/MH/day.
So, there's still some wiggle room for profitability, but it's very difficult to predict from here. I think difficulty should respond to price changes from here, which would essentially imply that for the time being, the profitability equation might not change much.
I would encourage discussion on this topic, since I can't really come up with any way to predict scrypt mining profitability going forward.
