The "low pressure when attached airflow wraps around curves" can be seen on many vector* diagrams of pressure distributions on cars.
(*force strength and direction)
Here's a classic one for the Mazda RX7 (but note it doesn't show undercar pressures - back then, no one much cared about that):
When examining the vectors, we can use the 'triangle of forces' to work out what is the horizontal (thrust/drag) component, and what is the vertical (lift/downforce) component.
From my book:
Mercedes in wind tunnel...
Developed vector force:
Yellow arrow shows thrust vector, and green arrow shows lift vector.
So, in pretty well all cars, the airflow at the top of the windscreen helps propel the car forward, and lift it off the ground. But, because the area where these pressures are applying is very small, both forces are also relatively small.
Now mentally apply the same triangle of forces to the vectors on the roof and hatch of the Honda:
You can see that the horizontal components are small, and that thrust and drag (including all the hatch) probably cancel each other out - that is, the roof and hatch are not causing any pressure drag.
Resolving the horizontal and vertical forces allows you to see how lift/downforce, and thrust/drag, are being created by the pressures acting on the body.
And a reminder: you can easily measure all these pressures for yourself by doing some simple on-road testing.