Start with N/Casting Cost % chance of copying an attacking creature.
The Absorb cost per Copy created should be kept (with possible scaling) but would fit better as 
.
Note: If there's a percent of failure, we don't need any limits, and the cost-per-copy would be a lot cheaper.
It makes sense with life because you are creating new life, like mitosis.
As it stands, this is the kind of shield that I would just drop in a mono-aether and chain.
The 
drain is so much that trying to keep it is not worth it, it would be simpler to just drop this shield out to copy everything for one turn and then play another one as soon as you get it.
I'd assume that if you don't have the quanta to copy the first creature, it'll be destroyed before the second creature attacks, not after all the creatures for that turn attack (like dissipation).
Also, in my suggestion "shield isn't destroyed after counters are gone and if you don't have enough quanta to copy," if you don't have enough quanta then a copy just won't be made.
