Originally posted by Tjompen1968 and with a 100% still subject.
True and false. Motion correction deletes the "blocks" where motion is detected. So if motion is detected in one block the other three are encoded, down to one block. Sort of the opposite of MPEG compression which encodes the blocks where motion is detected. So if motion is detected the full four blocks will not be used resulting in less detail resolution in the affected pixels.
View this seminar by Tanaka Kimio from the 2016 CP+. At about 20.:50 Mr. Tanaka discusses motion correction for Pixel Shift (Real Resolution on the Japanese menu). He then shows a Pixel Shift image of ice skaters with Pixel Shift MC off vs on. At about 24:35 Mr. Tanaka shows a "waterfall" shoot with a detail crop of the "white water" MC on vs off.
There were reports from PF members that attended CP+ or other shows in Paris or Canada of them taking Pixel Shift photos of the crowds at the show to demonstrate motion correction. The posters didn't say if a tripod was used or not.
The following interview in DC Watch (in Japanese use Google or a similar translation service) the Ricoh engineers discuss motion correction (about half way down):
http://dc.watch.impress.co.jp/docs/news/interview_dcm/20160322_748845.html
Flickr and Photobucket both seem to be down at the moment so I'll attach screen caps from the above video. Pixel Shift is referred to as Real Resolution (RSS) in the video.