What's wrong with buses includes that they take up a lot of space, have a high energy cost to run, tend to have a relatively low peak throughput, tend to be noisy and polluting, and you often have to wait a long time for the next one.
It's expensive to build these contraptions now but the cost is likely to come down over time. Once built, by energy cost per person they don't compete with "make people walk on a non-moving sidewalk" but do compete pretty well with most of the alternatives. You COULD build them in such a way that the cost of dealing with snow is minimal, it's just a matter of deciding on a suitable strategy and implementing that strategy. The dumbest, simplest strategy would be to add some sort of roof or canopy or shade structure such that snow and rain fall on either side of the conveyer belt. (Like, say, the conveyer belt entrance to Bally's Casino in Las Vegas: https://rlv.zcache.com.au/ballys_las_vegas_conveyor_belt_pos... )
Other options include digging a suitable drainage/collection area that snow can fall into as the belt turns or, yes, melting snow with a heater as it passes a grating. The heater option is only energy expensive while it's snowing but it's still likely cheaper than having guys with trucks clear it. But if the area just plain has too much snow for melting it to be practical you go with the roof option.
