The HSL/HSV color spaces are not considered "perceptually uniform" despsite being widely used. Is there a detailed reason for this?
This is just a physiological truth. Perceptually uniform colorimetric models are complex and there must be a strong justification to use them (why would you need perceptual uniformity, after all ?).
On the other hand, HLS/HVS provide a sufficient empirical decomposition in dominant color, lightness and colorfulness.
In a perceptually uniform colour space the perceived difference in colour of two colours is proportional to the difference of the values used to describe the two colours in the colour space. For colour spaces such as CIELAB and CIELUV this was the main reason for the creation of the colour space.
If two colours are represented in CIELAB with the values L1, a1, b1 and L2, a2, b2, then the perceived difference of the colours is intended to be the square root of ((L2-L1)^2 + (a2-a1)^2 + (b2-b1)^2). This means that if two colours are represented by values in CIELAB which are significantly different then the colours will be perceived as significantly different.
This is not true of HSL and HSV. Two colours such as H1=0, S1=100%, L1=100% and H2=359, S2=100%, L2=100% have greatly different values for H, but will appear nearly identical. Similarly H1=60, S1=100%, L1=1% and H2=240, S2=25%, L2=1% differ greatly in H and S values, but will appear nearly identical, being both nearly black.
HSL and HSV are useful and widely used because it is easier to tell from the H, S and L/V values what the colour is, but they are not perceptually uniform and should not be used as the basis to calculate the perceived difference of two colours. Colours should be converted from HSL/HSV to e.g. CIELAB to calculate the perceived difference.