There is a theory to that effect but I don't buy it. A cladistic analysis is only as good as its characters, and Haplocanthosaurus has many that don't match anything known in Diplodocoidea. Fowler and co. came out with a paper recently that claims Haplocanthosaurus is a juvenile of Suwassea or one of the diplodocid species, which is flat-out false in my book. Far too many changes would have to take place to make it an ontogenic stage of any known diplodocoid, particularly the lack or bifurcation in any of the spines.

Now I know that Rebbachisauridae have single neck spines but the body and hips are quite different from Haplocanthosaurus, not to mention the tail. Oddly enough the tails of rebbachisaurs like Nigersaurus are restored with whip-ends despite being outside of flagellicaudata. So then either this puts Haplo outside of diplodocoidea, or flagellicaudata is a pointless designation, or the rebbachisaur tails were really not whip-ended and more like Haplocanthosaurus.

The simple truth is that Rebbachisaurus doesn't fit comfortably anywhere in neosauropoda but clearly it is not more basal than the base of neosauropoda. The unsplit neural spines may or may not point to a proto-brachiosaur affinity (my best guess is no, since Atlasaurus and other basal brachiosaurs or brachiosaur cousins have some bifid neural spines, so the trait may have appeared and been lost several times in neosauropoda). However the sacrum of Haplo is definitely more like a brachiosaur than a diplodocid, and the sacrum is often the most diagnostic piece of a sauropod you can find. There's very little incidence of convergent evolution or secondary re-evolution/loss of traits in sauropod sacrum structure. If the sacrum were known for Volkheimeria, Lapparentosaurus, or ?Bothriospondylus? madagascariensis, it would probably resemble Haplocanthosaurus.

As for Amphicoelias... that's more of a mystery since A. fragillimus is gone (and may not have even been correctly illustrated) and A. altus is not very complete. However one phylogenetic analysis is not enough to give a final verdict (heck, every analysis done on titanosauria seemingly disagrees with all the others!) and cladistics is only as good as the data you use. It's very easy to have a computer program misinterpret convergence as immediate common descent, and many papers even enter characters that are non-existent or improperly described (this appears to be a much bigger problem with sauropod research than with any other group). Whatever Amphicoelias was, it definitely was not very much like Haplocanthosaurus.

Possibly. Opisthocoelicaudia may actually not be as fat as Ruyangosaurus or Puertasaurus. Restoring titanosaurs is tricky, even those with complete ribs. I suspect Alamosaurus was very wide, judging by its dorsals and how squat Rapetosaurus is. However some of its close cousins like Huabeisaurus and Jiangshanosaurus don't appear to be very wide. My best guess would be rib cage proportions similar to Futalognkosaurus, but maybe not as extreme as Puertasaurus.