In one of my recent posts I speculated about the "survivability" of elongated orthostats or pillars when carried by ice over long distances. It is obviously possible, as shown by some examples in the literature, but the stresses and strains involved make it much more likely that during transport in or under a glacier pillars will simply be snapped into smaller segments. I am no engineer, but common sense dictates that it is easier to carry -- without damage -- a block of stone with width:length dimensions of 1:2 than a block with dimensions 1:7. So that limits our options with respect to glacial transport. That's why I speculated on very large "super-erratics" being dragged from their places of origin and then being broken up either during or after transport in the ice.
Some of the Stonehenge bluestones (for example 69 and 70) do indeed appear to have pillar width:length ratios of 1:7 or thereabouts. So are those stones more likely to have been carried by human beings and their gadgets, or by glaciers? I'm not so sure about that. A number of people (including Aubrey Burl) have argued that it makes no sense for Neolithic stone-collecting parties to carry rough pillars weighing maybe 6-8 tonnes all the way from West Wales to Stonehenge, only to then cut them down to a pillar shape and a weight of 3-4 tonnes. That's not very ergonomic. Much more likely, one might argue, to shape them close to the place of origin, and then carry them on their sledges or in wicker baskets, or whatever.
Here are some of the transport devices suggested: top, the Len Saunders suggestion (a Cuban A-frame or V-shaped sledge). Middle: the Len Saunders suggestion for "underslung" raft transport. Below: the sledge used (with rollers) in the 1954 Atkinson towing experiment.