This revolves around recording performance tradeoffs. ACID attempts to write inbound audio to your hard drive as efficiently as possible. During multitrack recording, it does its best to transfer audio to your hard drive without missing anything. Other programs that write to a floating point format must, at minimum, convert the data to floating point format before writing out the data. This involves at least some processing overhead, which ACID is trying to avoid. Instead, ACID writes your raw audio data to the hard drive, and adds any input monitoring effects at a lower priority. If any audio has gaps, it will usually be your input monitored audio, not your recorded audio.