The by Litchi created DJI SDK log doesn't give any more firm insight to what went wrong than the Litchi .csv log.
Everything went good on the AC approach to the HP during the RTH phase ... the AC braked & came to a stand still just above the HP.
Looking into the differences between GPS & registered IMU velocity gives us the chart below, the marker is placed just where the braking is done & the AC have leveled out. The velocity differences is really close to zero as it should be, but then ... especially the northward speed direction starts to generate a big deviation. This usually is a proper sign of IMU problems, either due to a disturbed compass during power on or wrong IMU fusions of the sensor data.
View attachment 110803
Your crash movie clearly show the AC speed away in a southward direction ... that should show a tilt in the same direction. But here it's evident that the IMU misinterpret the sensor data & instead decides that the tilt direction is in a northward direction.
Here below in the pic the green bar show the AC heading ... & the blue one the tilt direction (tilt = combined angle of pitch & roll from the vertical axis. A leveled AC = 0 degrees).
View attachment 110804
I am thinking IMU failure with a possible crash sometime in the past or previous water damage. Maybe the OP can fill in those details ?
What software do you use to read the .dat file?
Showing the same in a chart (markers placed where the AC is in the pic above) show a large tilt inclination on par with using Sport mode according to the specs for a M2. The darker blue background area indicate a positional error & just there after, the AC behave unbalanced tilt inclination wise but also tilt direction wise. The most probable cause is a IMU failure ... but without a .DAT log everything ends here I'm afraid.
View attachment 110805
This incident had nothing to do with partly flying over water & ending up there ... just to avoid a lot of speculation about that.