One of the big reasons that a single cylinder engine cannot be balanced to no vibration is the variance in acceleration due to combustion and compression.

Imagine the typical two stroke engine cycle, starting with the combustion. At this point, the engine is at or near TDC. The combustion will cause a rapid acceleration of the piston downward due to the expansion of gasses. Those gasses are then vented out the exhaust port. This rapid acceleration downward also causes a rocking motion of the piston, due to the path taken by the connecting rod not being directly parallel to the cylinder bore's axis. As the piston returns back towards the top of the cylinder, the gasses now in the cylinder from the transfer port and the gasses pushed back into the cylinder by the reversion wave from the tuned pipe are compressed. That compression slows the piston down rapidly, or decelerates it. The difference is that the deceleration due to the compression, and the acceleration due to the combustion are quite different. That difference contributes to the vibration already occuring from piston rock and unbalanced reciprocating masses. If you could balance the reciprocating mass you would still have uneven accelerative forces causing vibration. One possible solution is to use more than one cylinder to try to balance the forces.

As a side note, an engine is not turning a "constant speed", but rather accelerating and decelerating rapidly enough for it to be hard to notice, and we give the average rotation a value relating to how many rotations the crank makes in a minute (RPM). The actual rotational speed (angular velocity) changes as a function of engine timing and where it is in the combustion process.