Its worth thinking about using the 'Sync + Pause' method instead (or the tag-triggered method if you have custom karts) - ?
I spent almost as much time on the lap-triggered barriers on 'Alphabet Avenue' as the letters themselves, yet still without a reliable result so used the 'Sync + Pause' method as an alternative and it seems to be fine.
To find ideal timings, playtest at a range of player speed / abilities with all features except the gates, then try playtesting again to tune the gate motors / pistons etc Sync and Pause settings. Remember to allow for player time through a gated section if you're also using an end gate, and include a moving / revealing bounce pad or jump to let them out just in case they still get trapped.
Perhaps the lap triggers are glitched or just temperamental - ? - I don't know. Another track of mine uses lap triggers to change track lighting / custom kart headlights etc and thats OK, but on repeat plays you can still notice some time variations even when consistently in 1st.