I think we’ve feature parity between old and new implementation.
Phosphors can reduce theirselves well at screen resolution without changing algorithm; slotmask/aperturegrille can do the same.
By nature, when they sould grow in width, (but have no room since there is no space left between them) they will simply blend themselves making the mask less evident and fading to the original color:
If one have an high dpi screen, he may choose to use the multiplier knob, while still staying at screen coordinates, to make room for phosphors and see they growing in width.
Being on 1080p, i can try that with nothing more than a tight gm mask, but I suppose that on higher dpi screens it would look much better.
Speaking of multipliers, to answer to your prevoius questions about slotmasks at core resolutions, the arguments are the same; if one have an high resolution display, it may prefer to use core coordinates and use the multiplier as a divider.
There are still many possibilities that need to be explored.
Eg: while you can still choose gm,gmx,rgb,rgbx,rbg,rbgx masks preset, you can freely draft your own.
The following picture is a bit extreme, because i’ve set a manual mask of just 1px, so that rgb phosphors are naturally blended into a single one; I’ve also set them to shrink and grow between 0.2 and 1.0, drawing them on screen coordinates, multiplied by a factor of 2, this looks very good on my 1080p and as a side bonus, adds more contrast to the picture without loosing color informations, because the contrast itself is gained by drawing less due to to the low min width (0.2).
Note that i activated an option to avoid intercell bleeding, that’s why you see a kind of grid:
