Is Loop Quantum Gravity trying to bring quantum mechanics from it's principles? Or is it taking ideas from quantum mechanics and merging them with GR to find a theory of quantum gravity?
LQG came from work on spin networks, which posed a simple question: Suppose you were unaware of space-time and looked at networks of interacting particles with spin, would space-time or something similarly arise naturally from their description?
Penrose only applied QM principles, he did not modify QM and did not bring in any GTR; rather he hoped that he could derive GTR-like results from his model. In this sense, LQG is much more natural than string theory which "generalises" QM. On the downside, not all the desired GTR-like results have yet been derived.
LQG in principle is trying to discuss gravity as a quantum field theory from a geometric perspective. It is not a "theory of everything" so it does not want to explain qm.
It evolved into a theory which is capable of discussing the evolution/interactions of a small (planck sized) chunk of space and also apply its principles to early universe physics (loop quantum cosmology, LQC).
If we have a working and renormalizable theory of quantum gravity then we can merge it with the standard model.
LQG is currently not trying to be such a theory, as quantizing gravity in itself is already hard enough of a problem!