Please start at Evolution.html

Conditions:
(a) Channelised flow starts on a rock surface.
(b) River incision slows progressively.
(c) Initially river incision, weathering and sliding produce a valley side slope at the angle of rest of the weathered material.
(d) The angle of rest is 45°.
(e) River incision produces basal steepening of the lower slope. When this steepening exceeds a certain critical amount shallow debris slides are triggered. These work their way up slope lowering the slope but maintaining its angle.
(f) At an early stage river incision slows to such an extent that creep is able to affect the whole valley side.
(g) Creep always removes more material from the top of the slope than the bottom.
(h) The rate of incision slows at each stage to such an extent that it allows more material to be removed from the top of the slope for each unit length the river cuts down. So, as the river cuts lower and lower, for each unit it cuts down more and more material can be removed from the valley side by creep.

Explanation of the simulation:
(a) The river is shown in blue, the rock in green and weathered fragments in yellow.
(b) The rock is divided up into rectangles to represent the fragments it weathers to. The angle of rest of the weathered material is given by the diagonal of the rectangles. The dimensions of the rectangles can be varied to give the angle of rest required.
(c) Each click on the manual simulation represents a longer period of time. The animated simulation is therefore misleading.
(d) The incision is shown in stages - but the lower the stage the longer period of time it represents.
(e) After the second stage creep is able to affect the whole valley side. To simulate the fact that more material is removed from the top of the valley side than the bottom if one unit is removed from the bottom of the valley side two units are removed above that and three above that etc to the top of the slope.
(f) If the valley side is to flatten at each stage then more material has to be removed at each stage. More material has to be removed from the top of the slope notwithstanding the gradient is gentler. This results in more material being moved from the entire slope. The only way this can happen is if downcutting slows by increasing amounts at each stage.
(g) The left and right hand edges of the simulation represent the mid points of the interfluves. Eventually the interfluves become crested and are lowered.
(h) Once the interfluves become crested the river continues to incise at an incredibly slow rate. Creep is still able to move more and more material at each lower stage. This means that the valley sides get very, very flat as incision almost slows to nothing.

Conclusion:
It seems very unlikely to me that this happens! Usually as river incision slows meandering produces a valley floor and valley sides then flatten as described in case 1 and case 2. Creep would very slowly flatten the constant slope of case 2.

The only circumstances that could produce the sequence of valley profiles shown in the simulation are a combination of rapid weathering, rapid sheet flow and initially very rapid incision that slows quickly. This might happen under periglacial conditions where frost action, solifluxion and meltwater work together to shape the landscape.