Somehow I really haven't succeeded in getting my point across. Thanks for all the practical advice, guys. That wasn't what I came looking for with this thread, as I've written a couple times, but I'm sure other readers will benefit from it. I agree with what all of you are saying, making deliberate choices about exposure is to be preferred over automation.
Yet, camera makers claim to have meters that are better than averaging meters. I say, if you want to refine metering beyond blindly following averaged readings, you quickly get to a point where you need to know if you're exposing for slide or negative film, and adjust accordingly. Which is what those of us that share my premise do when we're exposing deliberately. Matrix meters, however, don't.
I guess I could have phrased my whole question like this - why don't matrix meters do a standardized version of what a photographer with a spot meter does, something along these lines:
For slide or neg at night: If contrast is too hight to be reproduced completely, place highlights that cover an area larger than one metering point in zone VII. If that would mean that more than, say, half the image area falls below zone III, accept blown highlights and and expose for mid tones instead.
For negative: If contrast is too hight to be reproduced completely, expose so that no shadow area covering more than, say, three metering points falls below zone II.
Of course we who deliberately decide about our exposures would still sometimes want something different from that. But I imagine that in some critical situations, it would be better at what it's supposed to do than a metering system that is blind to its medium.
The answer to the question why why this isn't being done must be that there are practical obstacles to that level of sophistication, perhaps it would introduce too many new problems. Perhaps complex software is actually necessary for that. I wouldn't have thought so, but so it seems.