To be honest though, I don't understand your entry based on your explanation. In your circle, price is now way above the regression line, so its almost negating that line at that point, no? The downslopping blue trendline is really juicy, but at this point, your blue EMA is almost above the red one.
Now if I look at the yellow circle that I drew, now I can see how this regression line "reverses to confirm" because it tests from below (if this is what you mean by reverse to confirm), and the blue trendline is once again in play. Of course you never know if this opportunity comes up to give you this nice combination of upper trendline test and lower trendline regression from the opposite side, if I understand you correctly, but at least in this yellow area, the short seems to be looking better.
View attachment 344608
over the years you learn that trendline/regressions are just dispersions around a vector. Meaning a slight break is not that indicative of bias, repetitive breaks is more indicative. So the lines are a rough guestimate of the vector (direction) for the price. Its a summation of all the indications that creates the internal bias of direction which is the directional bias your entries should be. Meaning at any given time, you only trade with one bias, either looking only for longs or either looking only for shorts, unless your internal bias does a 180.
in the above picture notice the red line breaks but doesn't truly hold over time and the vector is down.