Reading a paper High-throughput purification of single compounds and libraries by Mathias Schaffrath 1, Erich von Roedern, Peter Hamley, Hans Ulrich Stilz, they divide purification in pharma into 3 phases:

1. Purifying Scaffolds - large amounts & small number of samples
2. Purifying Building Blocks - medium amounts & number of samples
3. Purifying final products - small amounts of large numbers of samples

They mention that for the 1st phase it makes sense to use Normal Phase while the rest uses Reverse Phase chromatography:
We consider that normal-phase chromatography is most suitable for the separation of larger amounts of material. Our standard stationary phase is ordinary flash chromatography grade irregular silica gel from Merck (Si60, 43-63um) prepacked in 57 x 300 mm plastic (FEPD) columns from Gotec. Smaller particle size of spherical silica gel does not seem to have any beneficial effects on separation quality. In comparison to reversed-phase silica gel, roughly five times higher loading can be achieved , and it is only a fraction of the cost to replace the stationary phase. In practice, we reuse the columns for 100-200 injections and wash between samples with pure methanol.
I was under impression that Normal Phase doesn't resolve organic compounds well and that's why pharma exclusively uses Reverse Phase.

The article was written in 2004 - is Normal Phase still used in HTP and is it it used in other parts of the drug discovery?