serch:
Sure, but whatever the writing effort compared to programming / art (which would admittedly be less), you're still talking about tripling the amount of writing work. Assuming that you've got a fixed time to spend on writing, wouldn't it be better to get five really well written options, than fifteen rushed ones?
Also, you can't view it as a totally disconnected enterprise. If the fifteen (for example) options are to make any difference, they need to have consequences. That means programming work. If you're talking about a game like ME, where results can include actions (e.g. pulling a gun), then it's also extra animation work. [assuming *pull gun on NPC* isn't something that happens every other dialogue]
This could be interesting, but you'd have to consider whether it's an efficient use of resources. For an RPG I'd have thought not. For a Hitmanlike game, it might be [fairly fixed character, variety of action-based consequences to dialogue, rather than intricately worded quest branches...]
I think with an RPG which allows a variety of characters, you'd lose too much in terms of expression, and still fail to cover most differences between characters - even with a lot of work.
Also, I'm not sure there aren't better solutions without needing to remove direct player dialogue choice. For instance, dialogue options can become available/unavailable based on character traits (as in Fallout/Arcanum...). E.g. for a sleezy character, the tasteful option could simply not be there. This kind of system could work based on many inputs - with character personality traits having an influence, as well as attributes etc.
Alternatively you can give a certain dialogue, but have the result based on a skill check (as in many RPGs). With a text based game it's simple for the player to imagine that his character said the line convincingly/unconvincingly based on the result.
For a voice-acted game, you could (in theory) record two (or more) versions of each line, and have the character say the one which corresponds to the skill check. So for example if you're making a forceful threat, but fail the check, you get the "nervous" voice-over line - and therefore don't expect things to go well.
These solutions also take quite a bit of extra development - in writing / scripting / voice acting etc.. They all give some extra variety without losing the Precise Control (TM)
over dialogue though - which is important IMO (for most RPGs, at least).
The main thing that is gained by having a more generic system (as in ME) is speed, from what I can see. That's also the "advantage" that's usually referred to. I just don't think this is a real gain. I'm sure that they will be able to design some situations where it makes good sense that you need to act quickly in a dialogue. However, that's what they'll be doing: setting up situations which suit their next-gen dialogue system.
It'd be quite possible to design a load of situations where quick action in dialogue is not necessary too. Certainly this would put certain constraints on design for a real time game. I don't think it's a harmful constraint though - much less harmful than denying me the right to choose what my character will say, at any rate.