The air has been accelerated above the wing, because at the end of the chord the air wants to merge again at the same point where it separated in front of the wing. Otherwise you would get a 'hole' in the air at the end of the chord on the upper side.
The profile of the speed of the air above the wing depends on viscosity of the air.
There is nothing wrong with using Bernoulli to explain or calculate lift. If you measure the velocity of airflow around a wing section, Bernoulli will accurately predict the pressure differential. What is wrong with the typical Bernoulli explanation is what I've bolded in your quote. The air doesn't meet at the same point, if it did that there wouldn't be enough of a velocity difference to account for the lift. The velocity of the air going over the top vastly out strips that of the air around the bottom and the air over the top
beats the air it separated from at the leading edge.
So to summarise, Bernoulli accurately predicts lift from the variation in velocity around a wing, but the "air over the top must meet the air across the bottom at the same point it separated" is wrong.
On the other hand, the typical "deflection" of air theory completely ignores the importance of the shape of the top surface of the wing in accelerating the air mass.
Newton and Bernoulli are both correct, they are both different ways of measuring the same thing. In simple terms lift is produced because the shape of the wing along with the angle of attack of the wing causes the air mass to be accelerated downward. Bernoulli accurately predicts pressure differential and therefore lift from the difference in velocities around the wing, Newton accurately predicts lift in terms of air being accelerated downward. So Bernoulli and Newton are both correct but
the typical explanations of lift using Bernoulli and Newton are both wrong in important aspects.