I think the most likely scenario for this to occur is if a normal warm-sector tornado were to pass over a surviving snowpack produced from a previous winter storm. The snowpack would only need to be deep enough to survive extended melting (several days) from being in the warm air prior to the tornado event, and it shouldn't have an adverse cooling affect on the boundary layer if return flow was strong enough. This has 'almost' happened a few times that I am aware of, but the snowpack did not survive the warm return flow and entirely melted long before the severe weather event.
For example, picture if today a strong trough suddenly dug into the Plains and started vigorous return flow of 60F-65F into southeast Kansas. Providing a narrow and sharp instability axis and extreme shear - as long as the snowpack survived the brief warmup, any tornadoes in SE Kansas and Missouri would be white with snow.
I've never seen the snowspout example - if I'm not mistaken though, the snowspout is a rotating column of fog on the lake rather than a column filled with snow?