As things are now shields hold energy, which directly translates to hp. When a shield gets depleted, either by damage or base power usage, a crew member will bring a battery which instantaneously replenishes shield strength. If a shield is right next to a power source, it will start getting resupplied in just a few seconds after initial damage, and is extremely likely to restrengthen before going down. If a shield is farther away from power supplies, then it is far more likely to go down before any batteries arrive.
Now obviously it is intended that putting any device closer to whatever fuels it will make it perform better, but I feel that in the case of shields, the difference between a shield that has a reactor right next to it and one that has a reactor like 3 blocks away was already too big, and it's especially the time lag between initial damage and the first battery arriving that does it.
I also believe that if double/triple batteries will be as they are in the current large reactor test version, this will exacerbate the issue.
So what I propose is that shields will be buffered: they will store the same (well, similar, numbers would have to be changed to not have this be a general buff or nerf) amount of energy they do now, but rather than shield hp being equivalent to the current amount of energy, they will convert energy into shield hp at a set rate while the shield is not at full strength, and do so immediately after taking damage.
This way placing reactors closer to shields will still be rewarded as their buffers will get resupplied faster, but the initial difference in virtual hp pool is not nearly as big.
Aditionally, buffer to shield conversion rate is an extra tool that may help to balance different sizes/varieties of shields, including in mods.