In Cecil’s recently published classic column, Is Mercator projection on maps a plot to reduce importance of Africa, S. America?, Cecil mentions that the National Geographic Society recently (the column being published in 1989 and the NGS switching over in 1998) changed their standard world map projection from the Mercator to the Robinson. In over 15 years since the column was published, the NGS swtiched over again, and now uses the Winkel Tripel projection instead, which appears to be a little more squashed than the Robinson is.
My mistake, I was half-reading the column. The projection the Society rejected in 1988 for the Robinson was the Van der Grinten, not the Mercator (which Cecil stated they last used in 1922).