You didn't use Paint, so your diagram is not so good.

My other problem with your diagram is in knowing where the tangent line is in the actual situation. Ideally, yes, we stand perpendicular to the surface, however, you are not going to be able to tell that the horizon is below this tangent line (which we can't see). The declination is sufficiently negligible (like less than 1/100th of a degree). So, there really is no difference between our two diagrams given such small angles.

Basically, the video does not prove what the original poster thinks it proves. He's making too many assumptions, one being that "they are smart enough to make a perfectly level road." They may be smart, but the precision of road making equipment could not be good enough to come to a good conclusion as to the shape of the Earth.

Another thing: knowing the speed of the car we can easily calculate the difference in forces on the car between the "perfectly flat RE" model and the "10-foot hill over 5 miles" FE model.

`The angle of incline of this hill is: arctan (10ft / 5mi) = 0.02 degrees. `

Now the difference between forces on the car (traveling at the same speed in each model) is RE_{model} = cos[*above angle*] *FE_{model}

I can tell you without finishing the calculation that the difference is negligible because the COS of this degree measure is 1.0000000 (which is waaaay more significant figures than is necessary).

So the forces on the car between the two models ARE identical. Because of this, one cannot come to a good conclusion. There could be a hill. The driver would not be able to tell the difference, though.