I suspect that glucose is trafficked from the endosome-lysosome system into the cytosol. It is then metabolised through the pentose-phosphate pathway, trafficked through to the Kreb's or converted to glycogen.
In newborns, glycogen-autophagy serves as a response to overcome hypoglycemia. In such a situation, overall effect must be towards increasing blood glucose levels. I was wondering whether glycolysis is muscle is obligatory for glucose resulting from glycogen-autophagy. If so, the physioloical relevance of skeletal muscle autophagy will be highly dependent on liver.
As you imply, ever since Bernard, we usually think of liver as the source of needed plasma glucose. I am not familiar with the special case of a newborn, but for skeletal muscle the dogma would say that cytosolic glucose is so rapidly converted to G6P that the cell is very unlikely to establish sufficient cytosolic glucose concentration to drive secretion through GLUT1 or GLUT4. Another issue is the total amount of glycogen available in muscle versus liver. Again, dogma would probably say we cannot supply glucose from muscle glycogen in sufficient quantity to sustain plasma glucose. If this generalization is false for newborns, I'd like to see the evidence. Interesting question.