Yes, I'd agree, achieving float reliably with a small latex balloon is non-trivial. Properly characterizing the balloon parameters you need is where I think most of the challenge comes from. And the smaller the balloon is, the less margin there is to avoid bursting. I haven't done this with balloons <300 g, but that is also due to usually targeting higher float altitudes (>15 km) with payload masses that require this size. For this size, you need to calculate and measure free-lift to within about 5 g.
While I agree with your explanations sentiment, it isn't technically correct. The energy doesn't come from the gas itself (it is really the lift gas displacing a more dense gas upward and how that changes gravitational potentials). The lifting force is generated from the density difference between your lift gas and the surrounding atmosphere. During the initial ascent, the lift gas density decreases with altitude as the atmospheric density also decreases. This expansion is permitted by the balloon envelope as the lift gas volume doesn't completely fill the envelope's max volume.
As the balloon expansion starts reaching the elastic stretch limit, the lift gas density can no longer continue to decrease as the balloon ascends. The lift gas volume is now constrained by the non-stretching latex envelope, which keeps its density constant. This constant density means that the lift gas will be less and less buoyant as the balloon continues to ascend.
If there is enough free lift in your system then it is possible to continue to ascend beyond this point of max envelope stretch—and results in the envelope failing (rip or burst). If your free lift is low enough, where this slight relative density change can effect the ascent enough to stop the ascent completely before then envelope fails, then it is possible to achieve float.