>> Well, the reflectivity of color MIP LCD is not very satisfactory. It is barely adequate, even for people like me who are fans. This is both because of the narrow-band RGB filtering and the inherent losses of the polarization-based switching method. Even the "white" state is discarding most polarizations of the ambient light, and then the darker colors are even blocking that.

Yes, that's right. A typical color LCD transmits only about 5-10% of the light for white because of all those factors.

>> My fantasy is having the reflectivity be at least as good as good white paper, and with deep contrast too.

That exactly was our benchmark for mirasol development. We used to measure best-in-class color prints for color gamut, brightness, contrast, etc.

mirasol did not use polarizers or RGB filters. An advanced architecture (that I was leading) also avoided RGB subpixels, something which very few alternative technologies can do [1].

>> It also needs to be brighter in practice than normal objects because, no matter what, it will have to overcome some glare from whatever protective glass and touch sensing layers there are over the actual display.

Yes.

Integrated touch-sensing helps significantly though.

There are also optical means that can nearly get rid of glare, if cost were not an issue. I have seen demo coatings that make the glass practically disappear -- we would repeatedly walk into it if it were used on a glass door.

-------

[1] Liquavista had Cyan-Magenta-Yellow subpixels vertically stacked. A new Eink architecture uses multiple colored pigments within the same cell but now needs sophisticated mechanisms to control them independently.