TA is voodoo, it NEVER works reliably. Economics of the process is what actually works.
BTW: in the hypothetical surrogate TA observation that temperature is lower in the winter than in the summer (in northern hemisphere) what is more important: the "TA" observation or the knowledge of orbital tilt and planet rotation around the star?
Ah fundamentals v TA. An argument as old as the hills. The challenge being that economics is not the dismal science, but rather no science at all. And there is no rational economic explanation for short term price perturbations.
Think of it this way. TA is the study of the psychology of price discovery. TA is asking the question - what will the crowd do when X happens? And how can you make the crowd do Y?