The "needle" is a pitot tube. The Atlas A was just a test vehicle for many later Atlas technologies, but also had a lot of testing stuff installed, which was not used in the final product. The Atlas A was no ICBM yet, the Atlas D was the first.
The Atlas F has no pitot tube or air data sensor at all. It's guidance only uses IMU and rate gyro measurements. The Atlas D (which was used in Mercury and other missions), had only rate gyros but no IMU - it was radio guided, a feature which made it very unpopular as ICBM: The launch control center had to be very close to the launch site and needed large vulnerable radar systems.
Atlas E and Atlas F replaced radio guidance by a IMU, which also improved accuracy. This way, the LCC could be build underground and required only a minimal set of antennas. Atlas E and Atlas F only differed by small features inside the rocket, for adapting to the different prelaunch activities and storage: The Atlas E was stored horizontally and completely unfueled, the Atlas F was stored vertically and partially fueled.
For fixing the trouble with Multistage2: Turn it into a spacecraft3 vessel - you have no multiple stages in the Atlas A, the Atlas A was a single stage vessel with two LR89-1 engines. On the Atlas B, the sustainer engine was added and the three engines became the MA-2 engine complex.
Otherwise, remember that the booster engines are handled independently by the multistage2 module: Once started, you can't control them until jettison.