True, but a DM isn't just a rules adjudicator and reader of boxed text--a great DM is a great storyteller as well, and understands the basics: character, plot and pacing.
In other words, roll the BBEG's save behind the screen, and instead of announcing to everyone the number and declare the fight over, tell the group that although the casting of the spell felt right, something has gone wrong--the villain isn't dead/unconscious/incapacitated as expected, just severely injured/weakened/slowed...and angry! Fight on!
This is an even bigger rules kludge than anything complained about above.
Doing that occasionally and when narratively interesting is fine. But if you do it too often, why bother having a set of rules to adjudicate the combat? If players don't know what their tools will do when used (at least the vast majority of the time), I think the shared narrative breaks down pretty quickly.
My experience suggests that your expectations of the limitations of gridless combat are overstated. YM probably V, though.
RC
My experience and everything science knows about memory would suggest that a human DM's ability to adequately track position and movement of 6+ figures in a tactical space with the number of variables in D&D without visual aids is actually very, very poor.
As long as the DM is far and away better at it than the group and has the trust of the group, this will never present as a problem. But if any one of the players is as good at it as the DM or better, there are bound to be many, many mismatches of understanding the tactical situation.
You can add as much tactical depth as you want to a gridless environment, but the limits of the human brain mean that you will, of necessity, lose precision. That's not a bad thing, per se. It just depends on the preferences of the group and the experience and capacities of the DM.
If you want to include meaningful positioning tactics, grid is simply going to be more straightforward for the inexperienced DM or player, and more accurate even for the veteran. That accuracy will be more important to some people and groups than others. Some players might not care. The DM will describe the situation in a way that doesn't match their image, but even if they get hosed a bit, they will know that errors occur in their favor, too. That's just part of the limits of relying on memory. Other players might care a lot.
This is a great theory but the quality vs quantity argument is meaningless in a game where the players determine the nature of most encounters. How can a DM plan out a "meaningful" combat when he/she doesn't know when or where the next combat takes place?
In order for combats to be fewer and more meaningful they would need to be dictated in advance. Part of the fun for me is not knowing how much combat the PC's try and get involved with until they try. Attempting to inject too much meaning into an encounter that might never take place is wasted prep time for me.
I was speaking specifically of the published adventure. In that case, the meaningfulness of the encounters, and the number of them between narrative milestones was more or less dictated for the DM unless he/she wanted to rebuild it.
This gets at what I was saying earlier. The system encourages tactically interesting combats that take a long time. For me, building a module to fit the 4e system would involve a small number of individual encounters that each have a narrative heft commensurate with their gameplay heft. However, I felt that the modules I played were not like that. Instead, they threw many encounters at the players, many of them mundane or irrelevant.
