Scott's Manipulate Casting Time skill is an interesting idea, but I agree that it would be very difficult to quantify. My feeling is that the skill should have a fixed TF and that casting it would reduce the TF of the next skill by twice the TF of the manipulate skill. Thus if you have a 1:00 skill you can take 2:00 off the casting of the next skill.
The problem then comes in chaining them. If by definition the casting of the first reduces the casting time of "the next skill", and you cast it twice consecutively, the second casting winds up with a negative casting time and there's no benefit to using it for some big spell down the road--your 1:00 version will take 2:00 off whatever skill is next, but you can't possibly chain it to do 5:00 of castings and get 10:00 benefit. The obvious solution would be to allow the application of the bonus to apply to the upcoming select skill--but then you wind up with the potential of "saving" the bonus for a very long time, and even the potential that bias would become a factor.
That last is significant. Suppose I cast the one minute 4@8 manipulate spell five times, because I have this really potent ten minute fire strike and I want to reduce the casting time to 0. Let's say I've got a 15@ bias and I'm in a 15@ world, my BRA is 2@5 and my SAL 2@5, so I'm looking at 65% chance of success on this, 3% chance to botch. I roll 45, 75, 65, 97, 12--three successes and two failures. It will reduce my casting time by six minutes. But before I can cast it, I get killed, and verse into a new world. I've just done five castings of this spell, three of which were successful--but in this world, the world bias is 0@, so my chance of success just fell to 50%, and one of those successes is now a failure, and one of those failures is now a botch.
I do allow skills to be partially cast and then completed later (Graeme has designed several of these, usually involving material components which are used in an extended prep and then destroyed in an instant completion, -10 for allowing the break). However, I don't roll the dice for success or failure until the skill is completed--at which time I do so based on the conditions (particularly bias) when and where the spell is completed. That means that if this other spell is going to affect a future designated spell, then its success or failure must be based on the conditions at the time the effect is used, not the conditions when the prep is done.
I also don't like the idea this creates, that you could cast those five prep spells in the safety of your room and discharge them a month later when you need the power for that fire strike.
Thus I would say you've still got the same problem: you have to have distinct spells for distinct reductions in casting time, because you can't really chain such a spell if each casting affects the next casting.
You could, though, have a 1:00, a 2:00, a 5:00, and so on, used to shorten the TFs of very complex rituals.
And as Scott says you could also use a temporal effects spell to deflate the time for the casting--but these are usually rather complicated themselves.
John, you often tell me that since I say Multiverser can do anything, it should be able to do it the way you want it to work. That does not follow. If I have an all-terrain vehicle and I say it can drive anywhere, that doesn't mean it can do so without fuel.
Multiverser can do anything in the sense that whatever your character wants to do, there is a way to do that, and a way to make it work mechanically. If your character wants to take a skill that previously took him three hours to cast and cast the same thing in three minutes, he can do that--Multiverser makes that possible. It makes it possible by saying that he just designed a new skill based on the example of the old one, he gets the bonus of the SAL of the old skill, and if he is successful he adds the new three minute version to his sheet, with whatever terms apply to it.
So he can cast the "same spell" in the sense that he can get the same effect in three minutes instead of three hours, but mechanically he has used a different skill.
That is, using exactly the same mechanics at all times, the game can produce any desired outcome.
--M. J. Young