So, reading over "Task Actions" on page 120, there are two mechanisms for reducing the time a success takes by 10%: succeed with a Measure of Success (MoS) of 10, or "rush the job" and take a -10 penalty. However, that means that choosing to "rush the job" just makes it harder to succeed, without actually reducing the time the a success takes at a given roll. Actually, from the following table (for a 100 minute task action with a standard difficulty of 50), all it does is make you fail faster in place of the slower successes. Since the Measure of Failure (MoF) determines the time for a failure, what would have been failures before a rush job actually fail slower.
Timeframe (min) 100 90 80 70 Modifier 0 10 20 30 Target 50 40 30 20 Roll -----Result/Time (min)------ 90 F/40 F/45 F/48 F/49 80 F/30 F/36 F/40 F/42 70 F/20 F/27 F/32 F/35 60 F/10 F/18 F/24 F/28 50 P/100 F/9 F/16 F/21 40 P/90 P/90 F/8 F/14 30 P/80 P/81 P/80 F/7 20 P/70 P/72 P/72 P/70 10 P/60 P/63 P/64 P/63
So, from this, the only reason to "rush the job" is to reduce the maximum time--i.e., if I know a bomb will blow up in 85 minutes, I better rush the job by 20% to ensure that I'm not around at that point. It will never actually reduce the time that success takes for a given roll. Put differently, "rush the job" just converts some of the slower successes into fast failures, but is otherwise slower in every case.
In my mind, rushing the job should reduce the amount of time a given roll takes, at least for success, and probably for failure. Maybe you can take up to a -30 modifer to reduce the timeframe by 60%. That would produce the following table for the same situation:
Timeframe (min) 100 80 60 40 Modifier 0 10 20 30 Target 50 40 30 20 Roll -----Result/Time (min)------ 90 F/40 F/40 F/36 F/28 80 F/30 F/32 F/30 F/24 70 F/20 F/24 F/24 F/20 60 F/10 F/16 F/18 F/16 50 P/100 F/8 F/12 F/12 40 P/90 P/80 F/6 F/8 30 P/80 P/72 P/60 F/4 20 P/70 P/64 P/54 P/40 10 P/60 P/56 P/48 P/36
In this case, if you rush the job and succeed, you will always be faster than if you rushed less. Some of the failures take a little more time, but nothing as dramatic as a "slow" 10 minute failure taking 28 minutes to fail with a -30 timeframe modifier and MoF of 10 in the first case. This gives motivation to "rush the job" in situations other than the ticking timebomb scenario. This also makes intuitive sense to me. For example, if I'm climbing a cliff, rushing up will certainly make me more likely to fail, but if I'm lucky and manage to succeed, it should take less time than it would have otherwise.
My other issue with timeframes is the concept of failure taking 10% of the timeframe per 10 full points of MoF. For one, that implies that a MoF of 9 or less fails instantly, which doesn't make much sense (this can be solved by simply saying that a MoF of 1-9 takes 10% of the time, 10-19 takes 20% of the time, and so on). Secondly, there seems to be plenty of task actions that should not fail faster--if I decide to search a room for 15 minutes, I'm not going to "fail quickly". I'll just work off the assumption that it's up to the GM to decide when the fast fail actually applies, but it would be nice if the rulebook made that explicit.