Temporal Revision Theory
This theory was developed to explain the Telepod Paradox, which it does well. Its effects on other accepted theory are presently too large to explore, so it's being presented here as an adjunct.
Everytime someone time travels, the timeline changes. However, how is the new timeline formed? One could guess that after (in time error) every time travel event, the whole of the timeline runs through, from beginning to end.
Let's say that before the events of Chrono Trigger, the last time travel event was at the Ocean Palace incident. I propose that after Lavos flings everyone at the Mammon Machine throughout time, time runs from beginning to end with Melchior, Magus, and Balthazar in their respective places and times. This is the timeline that exists prior to the telepod experiment. However, the telepod experiment must exist in the timeline, along with everything that follows. Therefore, normal time in 1000 AD keeps going after Marle gets sent back to 600 AD. This results in Crono going back, which results in Lucca going back. Since no more time travel occurs in the timeline, the timeline runs to conclusion and starts again, with Marle, Crono, and Lucca coming out at the appropriate times, as well as all of the TTI time travelers. In this revision of the timeline, plot happens, the party leaves back for 1000 AD, and the Guardia line is never interrupted, plausibly allowing the Telepod experiment to run, which results in Crono, Marle, and Lucca never being seen again, causing shenanigans in the further future until the end of time. And so on.
The key difference from established theory is that there is no intermediate timeline between Marle time travelling and Crono time travelling. This would fundamentally modify the idea of time error, as time error would be not be established in units of time, but in indivisible, ordinal revisions. The first timeline would be have time error of X. The revision caused by the time travel there would have time error of X+1. There would be nothing in between. I suspect that this idea is far enough removed from the concept of time error, that calling it time error would be an error in itself. However, terminology is tangential to the concept.