Only for testing the codec. Otherwise such footage could be used only for taking still shots from video.
Some users reported how their settings managed to survive heavy shakes, shooting from running car, etc. If shakes are too strong or car speed too high, the difference between consecutive frames will be large. However, blur will be high as well so there will be no detail within particular frame (due to 1/50s or 1/60s shutter speed). I frames will be relatively small (due to blurring) and I suppose that P frames shoudn't be much smaller than I frames (probably they will, but not several times smaller). However, if shooting very detailed chart, "micro-shake" can seriously change the pattern on CMOS sensor while retaining relatively sharp image. This could be problematic. Static shots on a tripod should produce high I frames and relatively low P frames. At least this should happen in theory.
Some users reported how their settings managed to survive heavy shakes, shooting from running car, etc. If shakes are too strong or car speed too high, the difference between consecutive frames will be large. However, blur will be high as well so there will be no detail within particular frame (due to 1/50s or 1/60s shutter speed). I frames will be relatively small (due to blurring) and I suppose that P frames shoudn't be much smaller than I frames (probably they will, but not several times smaller). However, if shooting very detailed chart, "micro-shake" can seriously change the pattern on CMOS sensor while retaining relatively sharp image. This could be problematic. Static shots on a tripod should produce high I frames and relatively low P frames. At least this should happen in theory.