What the XRGB-3 (and earlier models) do is still technically deinterlacing. It's taking a 480i input signal and outputting a 480p signal.
It's not creating a fully detailed frame though. I believe it's better to keep line doubling and deinterlacing separate, because there are important differences. It's like the difference between component and RGB only more drastic. You can say that component has the same amount of information at the end display, but the fact is you lost unique pixel data between neighboring pixels. Similarly, you can say line doubling 480i = 480p, but the fact is you lost unique information every other line at the end frame.