((Sorry for double-post.))
Question about manips: in the new system, can you explain how one exactly uses a manip? Do you just input variables (heat to X degrees y metres away) and let 'er rip? If yes, could one 'prepare' a manip action by putting in the variables but not hitting 'execute' until needed? If yes, could this be used to regain reaction shots in a way, but only for a single prepared thing? For example, you prepare the input/calculations for putting up a forcewall, but wait with hitting execute until needed. And if you then need to do something else entirely, you first have to delete the prepared variables, wasting time.
I think it can be reasoned this way:
It is confirmed that amps are not "speed of thought/instanteous" (since an Arbiter can dodge/kill the amper before the space-sorcery is done).
Modern manipulators are ampers-in-a-can, basically; inputting the data does nothing except giving the brain-pilot target parameters.
Thus, while input methods can theoretically be streamlined, reaction shots are still largely limited due to the same tiny-but-not-insignificant amp "casting time" - i.e. if ampers can do 'reaction shots', then manips also can, and vice versa, if ampers cannot (in certain circumstances) then manips cannot either. Really, for all the changes it made, the new system really removed a lot of mechanical differences between the amps and manips (giving even more importance to the defining characteristics of each type).
Of course, this is pure speculation, and I already can see some small loopholes/backdoors Piecewise left for himself which can be used to reason a wholy different picture.