Inferring the timing of porphyroblast growth in the absence of continuity between inclusion trails and matrix foliations: can it be reliably done?