The only real difficulty can be that if your ascent is too rapid then the atmosphere thins out too quickly. Your engine doesn't have time to accelerate you to the new required speed and you start to stall out. Try flying an XR2 with maxed out ISP and don't use the scram jets. When you get to about 50 km fly more or less straight until you begin to heat up and then regulate your temperature by pitching up slightly. Assuming you keep your temperature up but in the safe zone you will find that you continue to slowly gain altitude and velocity without any stalling problems at all until you pass the Karman line and break orbit.
That was how I started testing a lot of my theories but I found that I just had to rewrite certain routines to produce more genuine 'Traveller' results. Lifters aren't just the hover engine firing downward (although in that clip they may sound like it) with a given amount of force. It's actually a force that is vectored directly against gravity regardless of the orientation of the ship.
That reminded that I did some small amount of coding. Basically my idea was similar that if you could control gravity, you could do what you said above, have an engine or lifter pointed against the local gravity vector and negate it. Then your use m-drive to establish a thrust vector.
The reason I did that was I was playing around with turning it off and on. For example if you are approaching a planet with it on then all travel is in straight lines. However if you wanted to curve around the planet, you get your initial velocity going in the right direction (usually tangent to the lowest altitude you want to be at) then turn off the lifter and let the local gravity turn your path into a curve. Using gravity to give you some free thrust beyond what your m-drive can provide.
By turning the lifter on and off at various times you can do some interesting maneuvers in near orbit.
The only problem I have with it is that it only just matches gravity in a straight down direction while lifters clearly have the ability to overcome gravity (allowing them to take off straight up) and vector the gravity (allowing them to move forward). While I've got the math all worked out (a lifter is capable of altering the vector of up to about 60% of the gravity acting on the body, meaning it can vector the full 60% up and accelerate at about 2 m/s since the 60% has to overcome the remaining 40% giving you and upwards acceleration of 20% gravity) I've been trying to work out a control system that wouldn't be overly difficult.
If you code up your craft using the API in a C++ dll then you can add a custom thruster. The API will give you a handle that you can store. You then use the handle to access the other API functions that allow you to point the thruster in the direction you computed.
I do this in order to properly simulate the different attitude control modes of the Mercury and Gemini spacecraft. Constant Rate, Pulse*, etc.
Incidentally, the 'true' aerodynamics of the scout ship are horrible. I always thought it would be possible to fly it in an atmosphere using the body as a lifting wing by angling it in flight, but it turns out not really. Even with the addition of some stabilizers it just isn't flyable without aerodynamics that have nothing to do with its shape or lifters.
Interesting I didn't get that far and aerodynamics is not my strong point. Simulating Ship Systems and Operations is where was I was good at. All I will say that Gravitics cures all ills when it comes to performance and aerodynamics. It not totally a free lunch and certainly not point and shoot especially at low-G. But compared to flying a Gemini or even a Delta Flyer it is a true hot-rod.
=====================================
For those who are interested here is the deal with attitude and transition controls.
Most people think of thrust control as something you turn on for so long, and then turn off after which you are rotating or in the case of translating moving at some rate of speed.
Well with electronics (and computers) you are not limited to that. There are several control modes that were used historically that made orienting and translating a spacecraft a lot easier for a pilot.
First all this relates to how the thrusters respond to the joysticks. There is typically one joystick for attitude control (roll, pitch, and yaw) and another for translation (Fore and aft, left and right, up and down).
One is direct, the further you push your joystick out the more thrust you put out. The thruster will keep operating for as long as the joystick it extended. This is what most people think happens.
The second is rate control. The joystick is extended and amount of extension determines the RATE at which the craft rotates or move. The rate at maximum extension is designed into the system. For example 10 degress/ second or 1 m/sec. The Mercury Spacecraft had this mode. When you release the joystick it goes back to center and the craft will deceleration to zero velocity. This mode was used only for attitude control in Mercury, Gemini, and Apollo.
A third is pulse control. The joystick is extended and when it reaches a breakaway point, the thruster will fire a short pulse (like 1 second). The craft will continue to translate or rotate after the joystick is centered. Controlling your speed of translation or rotation mean pushing the joystick out then in X times. Then when you want to stop doing the same thing X times in the reserve direction. This was used for attitude and translation control on the Gemini. Of the modes discussed so far this was my favorite and allow for very precise control over your speed when docking or maneuvering.
There are others more specialized modes for example keeping the craft level with the horizon. In the Mercury spacecraft this was done through relay logic tied to gyroscopes on the Gemini and Apollo through an on-board computer monitoring a gyroscope.
If you heard the phrase Gimble lock (for example in Apollo 13 movie) it because the craft uses a 3 axis gyroscope. There is are two orientation that will cause the mechanics for two axis to line up which will have the bad of effect of causing the gyroscope to lose it zero reference point among other things. This can be fixed by adding a fourth axis slaved 90 degrees to the innermost axis.
The Mercury spacecraft had a three axis gyroscope so was prone to gimble lock.
The Gemini spacecraft had a four axis gyroscope and didn't have the issue.
But Apollo also had a three axis gyroscope. This was a result of the fact that the Gemini and Apollo were developed independently.