$ \def \J {\Delta} \def \half {\frac{1}{2}} \def \kwart {\frac{1}{4}} \def \hieruit {\quad \Longrightarrow \quad} \def \slechts {\quad \Longleftrightarrow \quad} \def \norm {\frac{1}{\sigma \sqrt{2\pi}} \; } \def \EN {\quad \mbox{and} \quad} \def \OF {\quad \mbox{or} \quad} \def \wit {\quad \mbox{;} \quad} \def \spekhaken {\iint} \newcommand{\dq}[2]{\displaystyle \frac{\partial #1}{\partial #2}} \newcommand{\oq}[2]{\partial #1 / \partial #2} \newcommand{\qq}[3]{\frac{\partial^2 #1}{{\partial #2}{\partial #3}}} \def \erf {\operatorname{Erf}} $

Extreme Cases

When looking for a sensible meaning of the determinant $\sigma_{xx}\sigma_{yy} - \sigma_{xy}^2$, it can be proved in the first place that: $$ \sigma_{xx}\sigma_{yy} - \sigma_{xy}^2 \ge 0 $$ Which is a direct consequence of Schwartz inequality: $$ (\vec{x}\cdot\vec{y})^2 \le (\vec{x}\cdot\vec{x})(\vec{y}\cdot\vec{y}) \slechts \left(\sum_k w_k x_k y_k\right)^2 \le \left(\sum_k w_k x_k^2\right) \left(\sum_k w_k y_k^2\right) $$ Remember the Ellipse of Inertia: $$ \sigma_{yy} (x - \mu_x)^2 - 2 \sigma_{xy} (x - \mu_x)(y - \mu_y) + \sigma_{xx} (y - \mu_y)^2 = \sigma_{xx}\sigma_{yy} - \sigma_{xy}^2 $$ For $\sigma_{xx}\sigma_{yy} - \sigma_{xy}^2 = 0$ it reduces to: $$ \sigma_{yy} (x - \mu_x)^2 \pm 2 \sqrt{\sigma_{xx}\sigma_{yy}} (x - \mu_x)(y - \mu_y) + \sigma_{xx} (y - \mu_y)^2 = 0 $$ $$ \slechts \sqrt{\sigma_{yy}}(x-\mu_x)\pm\sqrt{\sigma_{xx}}(y-\mu_y) = 0 $$ Thus the ellipse is degenerated to a straight line segment.
Next consider the following fact: $$ \left(\frac{\sigma_{xx}-\sigma_{yy}}{2}\right)^2 + \sigma_{xy}^2 \ge 0 $$ On the other hand: $$ \left(\frac{\sigma_{xx}-\sigma_{yy}}{2}\right)^2 + \sigma_{xy}^2 = \left(\frac{\sigma_{xx}+\sigma_{yy}}{2}\right)^2 - \left(\sigma_{xx}\sigma_{yy} - \sigma_{xy}^2\right) \ge 0 $$ $$ \hieruit 0 \le \sigma_{xx}\sigma_{yy} - \sigma_{xy}^2 \le \left(\frac{\sigma_{xx}+\sigma_{yy}}{2}\right)^2 $$ And the maximum is obtained for: $$ \left(\frac{\sigma_{xx}-\sigma_{yy}}{2}\right)^2 + \sigma_{xy}^2 = 0 \slechts \sigma_{xx} = \sigma_{yy} \EN \sigma_{xy} = 0 $$ $$ \slechts \left(\frac{\sigma_{xx}+\sigma_{yy}}{2}\right)^2 = \sigma_{xx}^2 = \sigma_{xx}\sigma_{yy} - \sigma_{xy}^2 $$ Remember the Ellipse of Inertia: $$ \sigma_{yy} (x - \mu_x)^2 - 2 \sigma_{xy} (x - \mu_x)(y - \mu_y) + \sigma_{xx} (y - \mu_y)^2 = \sigma_{xx}\sigma_{yy} - \sigma_{xy}^2 $$ Which is now reduced to a circle: $$ \sigma_{xx} (x - \mu_x)^2 + \sigma_{xx} (y - \mu_y)^2 = \sigma_{xx}^2 \slechts (x - \mu_x)^2 + (y - \mu_y)^2 = \sigma_{xx} $$ Summarizing:
Determinant $ = $ minimum then Ellipse of Inertia $ = $ Straight Line Segment.
Determinant $ = $ maximum then Ellipse of Inertia $ = $ Circle.
Thus the determinant $\sigma_{xx}\sigma_{yy} - \sigma_{xy}^2$, eventually divided by the trace
$(\sigma_{xx}+\sigma_{yy})/2$, is a $0 \le $ measure $ \le 1$ for the "roundness" of the ellipse.