I believe there is texture compression built in to DirectX in the form of DXTC, which if implemented in Blitz3d would allow far more textures in a scene. I think the compression rate averages something like 1/4 of the original bitmap size.
If this, and a third set of UV co-ordinates, were added to the Blitz3d file format, it would allow for some very high quality scenes, as you could use very large lightmaps and have another layer for detail (e.g. dirt maps).
So - base texture on UV 0, lightmap on UV 1, dirt/decals on UV 2.
Of course I might be totally wrong.
If this, and a third set of UV co-ordinates, were added to the Blitz3d file format, it would allow for some very high quality scenes, as you could use very large lightmaps and have another layer for detail (e.g. dirt maps).
So - base texture on UV 0, lightmap on UV 1, dirt/decals on UV 2.
Of course I might be totally wrong.