Here is a simple but partial explaination that I prefer. The energy of the airflow is proportional to the square of the air speed. If the air speed doubles the energy is four times greater. If the airspeed tripples, the energy is nine times greater, etc. Therefore the energy in the flow is much, much smaller at model speeds. In addition, the energy is proportional to the volume of air involved at standard conditions. Model flight involves a smaller volume of air than full scale. Air clings to the surface of the object passing through it. Therefore, a boundary layer is formed between the air clinging to the surface and the air passing by the surface. The more energy there is in the flow the easer the air is brought up to speed as one traverses the boundary layer and therefore the thinner the boundary layer. So large fast objects have thin boundary layers and small, slow objects have thicker boundary layers. The boundary layer adds a virtual thickness to the object. If the boundary layer is laminar then the boundary layer is thinnest. When the boundary layer becomes turbulent it thickens and when the boundary layer seperates it becomes thickest. The drag is related to the object dimensions plus the boundary layer thickness. The lift is related to the angle of attack. Since the energy available at model sizes and speeds is less than at full scale, the energy available to keep the flow from seperating is lower for an airfoil operating at model sizes and speeds. This means that an airfoil operating at full scale will stall at a higher angle of attack than the same airfoil operating as a model. Because the airfoil operating as a model stalls sooner, it can not generate as high a lift as it could have under full scale scale conditions.

Reynolds number is just a convenient way of comparing flow conditions. When the reynolds numbers are the same, the flow conditions will be similar. The larger the difference in reynolds number between two flow situations, the greater the difference will be in the flow. Full scale is associated with high reynolds numbers and models are associated with small reynolds numbers. Insects are associated with smaller reynolds numbers than models. The low reynolds numbers associated with insects result in such poor airfoil performance that insects couldn't fly if they relied on the same flow mechanisms as models, most birds and full scale. So, insects use a clap-twist- snap mode of flapping that produces a stronger partial vacuum above their wings and a more powerful down wash under their wings. This mode is only possible in small sizes because of structural advantages at small sizes.