Step 1:
Created two five-by-five bridges, each surrounded by a five wide moat of smoothed floor and a wall. The bridge on the left side is a retracting bridge. The bridge on the right side is a drawbridge raising to the north. Both bridges are linked to the lever to the north.
Step 2:
Temporary platforms were constructed above each bridge, covered with bone stacks, and then deconstructed. This left both bridges covered with obsidian blocks and stacks of bones. A few of the bones did evaporate during he process due to the experiment taking place aboveground.
Step 3:
The lever was pulled.
Blocks and bones were flung off both bridges. Objects flung off the retracting bridge on the left moved a few squares in a random direction, the furthest of which landed 3 squares from the bridge. There did not seem to be a consistent difference in distance of travel between the blocks and the bones. Objects flung off the raising bridge on the right also appeared to travel in a random direction, but the wall created by the raising bridge prevented any blocks from landing on the north side of the bridge. A few of the blocks thrown by the raising bridge flew 5 squares away to land against the containing wall, but the majority traveled no further than those thrown by the retracting bridge.
Step 4:
The lever was pulled again to lower/extend the bridges.
Objects under the retracting bridge on the left which were still within the footprint of the bridge were thrown a few squares. The direction was random, and the distance appeared to be comparable to that when the bridge raised. Objects already outside the footprint of the bridge were unaffected.
Objects under the lowering drawbridge were disintegrated instantly. Objects outside the footprint were unaffected.
Conclusion:
Both retracting and raising bridges throw objects resting on them when retracting. The direction of travel appears to be random in both cases, but the wall created by the raising drawbridge will tend to block objects from moving off the bridge in that direction.
There does not appear to be a consistent difference in the distance which objects travel after being flung by the raising bridge versus those flung by the retracting bridge. The furthest objects thrown by the raising bridge traveled further than the furthest objects thrown by the retracting bridge, but it is not clear from this trial if this is random or a consistent difference. Further trials will be needed.
Retracting bridges, when extending, produce an object-flinging effect similar to that created when they retract. Lowering drawbridges, of course, do not.