No One Can Explain Why Planes Stay in the Air
If no one can explain why planes stay in the air, how does mankind build better airplanes?  The answer seems to be that even the academic and industrial leaders tend to focus on making incremental improvements on proven designs.  For example, the BWB design is the addition of high-aspect-ratio wings to the sides of a flying wing fuselage (with a blended surface merger of the sections).
 
We can do much better!  The below heuristics are founded in physics, but they sufficiently combine concepts to enable to human mind to grasp and combine with other concepts in the innovation process.  They were validated with computational fluid dynamic simulations and are updated as greater insight is attained.  Several papers and patents use these heuristics to identify paths to significantly increase what is possible in air and ground transit.
​

Heuristics (rules of thumb) on Generating Aerodynamic Lift

For Airfoils (2D cross-section)

1. Air’s velocity impacting a surface generates higher Pressure (“P”).
2. Air’s velocity diverging from a surface generates lower P.
3. Air expands at speed of sound from high-to-low P (except boundary layer).
4. Air’s velocity bends toward lower P and away from higher P.
5. A surface’s L/D is the normalized* surface integral of P tan^-1(α) on the surface. 
6. Shear forces are negligible at Cd > 0.01; Cshear drag < 0.001 for smooth surface.

Sources (Propulsion Sources)

1. ↑ Source power ↑ Lift at trailing edge of Lift-Spans; when Lift-Span pitch angle   < 0° from horizontal, ↑ Source power ↑  L/D as limited by Operating Point. 
2. For L/D > 40, induced thrust is needed; induced thrust subtracts from Drag.
3. For the highest L/D: the L-Span, Nose, and (optional) wing need to have pitches coordinated with L-Span interaction with a Source.
4. High induced Drag on surfaces at P > 3° should be eliminated, as possible, by transforming P to near-FS P or transforming surface cause induced thrust. 
5. A mid-chord Source can propagate lower P forward to increase L/D.
6. Optimal L-Span pitches decrease with increasing Source power.

For Wings (i.e. 3D airfoils)

1. Lift P is lost over side edges for a wing relative to an airfoil.
2. Use camber, fences, and distributed Sources to compensate for lost Lift P.
3. High AR wings may be added to low AR lifting bodies to increase L/D.

Toward ↑ ↑ L/D and A Superior Optimal Aircraft Design

1. Distributed propulsion technology bridges otherwise separate sectors of VTOL, BWB, tubular Fuselage, and HAPS/HALE aircraft into a single optimization space with a superior optimum (i.e., ↑  L/D ).
2. Use heuristics to create base case design, optimize with CFD.

      * Normalized per unit vector perpendicular to surface; lift is vertical component, and drag is horizontal component.