User blog comment:LittlePeng9/SCG(1)/@comment-5529393-20130313042828/@comment-5529393-20130313073720

One thing we could try to do is embed labelled or unlabelled trees into subcubic graphs, a la Goucher, thus showing that SCG(1) is larger than tree(n) or TREE(n) for large n (the full version of tree(n), not the subcubic version I've been using above.)  This may be not be possible for SCG(1);  perhaps it could be done for SCG(2)?