I noticed an issue with negative pulse propogation across repeaters (where a short duration of powered off state is propogated along repeaters.) It seems repeaters will:

1. No longer register/accept a negative pulse if the duration is less than the number of ticks the repeaters are set to.

2. If a negative pulse is the same number of ticks as the repeater setting, the break in powered state (signal) will propogate until just before the final repeater, whereupon it disappears, failing to depower whatever the chain was leading to.

3. If a negative pulse is larger than the repeater setting, it propogates normally.

The picture illustrates case #2, where the redstone becomes unpowered for 3 ticks, and the repeaters are set to 3 ticks. The magenta lines indicate where the pulse disappears.