And a bit further on, more of the same sort of techniques, collapse-ment of loops away from corners where we don't need them any more...
...leading to more evenly spaced, larger density polys overall...
Next step would be to notice that we don't need that polygon density we have at the top of the recess there to describe the curvature towards the back, so it could be further collapsed by turning a lot of those edges into the centre so that only a minimal number ever reach the back side, though ideally we should bear in mind the number of edges it needs to meet itself at the lower side of the recess when the loop around the mesh is completed.
..and because that whole thing is flat both top and bottom, there are endless places to solve anything we need to, which is partly why I haven't so far felt the need to resolve those complex poles I allowed to creep in there ! I would resolve it later though, as it is easy enough to do so with judicious use of Spin Edge, or just by again, turning them centre-wards instead, like this...
I'm away over xmas so can't show you more than that until I am back...
Anyway, Happy Holidays
CBR