There are three kinds of shielding, so it makes sense to look at how this could work:

1. Blocking with a wooden shield. Having an actual hand-held shield boosts blocking, which amounts to a percentage of mitigation of incoming damage. In this ignoring 35% of mitigation is easy to figure out. If an opponent is blocking 50% of incoming damage, the Lotus Fan effect is either multiplicative (effectiveness reduced to 32.5% mitigation) or subtractive (effectiveness reduced to 15% effectiveness). The former would be better if this is the kind of "shield" we are talking about.

2. Blocking with a magic effect. This is similar to the last type but doesn't require a shield. Like Magma Armor. Same kind of calculation.

3. Armor boosting shields. Similar calculation to the previous examples but applied to the increase in physical/spell resistance.

4. Damage absorbing shields. This includes effects from Brawler, Shielded Assault, Sun Shield and morphs, Obsidian Shield and morphs, etc. This effects of 35% less effectiveness for these types of shields would work differently. One way it could work would be to dissipate 35% of the absorbing shield then apply the damage from the Lotus Fan (which would be absorbed by the remaining shield). Another would be to ask how much damage the ability would do if 35% of the shield was taken away, apply that damage, and leave the shield intact. The latter would be practically useless except for situations where the shield was already severely weakened.

As for the first version of negating absorbing shields, the 35% dissipation could either happen only on the initial "hit" or could happen on each proc of the DoT. If it happens on each proc, the 35% would be based on whatever was left of the shield. Of course, the caster could stack different shields or recast the same one, probably with little effort because reduced cost or high regen. But having an ability that can also be re-applied to eat through such absorbing shields would change the strategy and flow of combat.

My vote for something to try out would be for affecting damage absorbing shields, the version that continually eats away at the shield. Whether 35% per proc is a balanced number (not too high or low) is an open question. Maybe 35% is OK if it applies only the first hit but some lower number is more appropriate if it procs on the DoT as well. Maybe it depends on whether the snare is kept. But whatever the formulation, it would add something interesting to PvP.