No bug here, just normal behavior.
When I apply this absurd HT curve to an image and completely blow it out, making most pixels white
Not really. An histogram transformation in PixInsight will never clip a single pixel unless you explicitly define a shadows clipping point > 0 or a highlights clipping point < 1. The curve you've applied in this example does not clip at the highlights. Bright pixels are not white on the resulting image (except those that already were white before applying the transformation), although they may
seem white due to limited gamut of the screen.
why isn't there a big peak on the right of the output histogram?
Because you have defined a shadows clipping point > 0, which is setting a 1.0267% of the total pixels to zero. Thus you have a prominent peak at the left side of the histogram, while you don't have any white pixels. The rest of pixel values != 0 are in fact smoothly distributed along the available numeric range. The peak at zero value is (by far) the maximum histogram peak after the transformation.
Histograms are always represented
normalized by the HistogramTransformation tool. This means that the vertical dimension of the histogram always corresponds to the maximum histogram peak. You can zoom the histogram vertically if you want to inspect the 'hill' around 0.95 on the horizontal axis, which corresponds to the initial histogram peak.
