Here are my readings of and suggestions to respond to the reviewers' comments:

\paragraph{Uniform dielectric coefficient.} Perhaps the best defense is to refer to the MSA papers describing concentrated bulk electrolytes (Lesser) and Kienke's measurements of dielectric coefficients in such solutions. The measurements indicate moderate reductions of eps (low frequency limit) for 5 M salts, typically to about 40. If you scale the enthalpy of hydration using Born (with an empirical radius to get the eps=80 value right), the change from eps=80 to eps=40 changes the hydration energy about 20 kT for Ca2+ (I have not done the numbers, only estimated). Thus it is not 100 kT.

How can you get away with this not considered? Well, the Born and the screening energies both scale inversely with an ionic radius. For each one you need a slightly different ionic radius, but the difference tends to be small. For instancs, for NaCl, the difference is so small that the enthalpies of hydration (at infinite dilution) and of screening in the solid salt crystal (absolutely no water there) are only 1.5 kT different. For CaCl2 the situation is a bit different as things get a little hot when you throw the salt into water, but I cannot recall a measurement with Ca formate, which is more relevant for your system. My best guess is that, yes, eps is less in 5 M solution, but the changes of hydration and screening enthalpies (that is, the changes due to eps) largely compensate because both energies scale similarly with eps, but the changes have opposite signs.

Regarding the effect of low eps in the pore wall, you can refer to our MC work: if the pore is pre-loaded with a fixed charge, the pore with LOW eps in the wall will attract MORE Ca into the pore. That's fact. The fact is the opposite of the behavior of pores not having a net fixed charge, which has been described by Chung's group. The low-eps pore wall does not act as a barrier (reducing local carrier density), but works the other way. The Chung way of making small conductance does not work in a region with fixed charge. You have fixed charge all over the pore.

I have the impression that this referee might be a benevolent former member of the Chung group, perhaps Toby.

As an afterthought: the idea of an electrostatic barrier limiting ion flow in Ca channels is questionable on experimental grounds. We need different D's for each monovalent and divalent ion species to match the experimental conductance (after concentration has been dealt with in the DFT). It has never been demonstarted how an electrostatic barrier (due to low eps in the pore wall) can distinguish ions beyond their net charge (e.g., K from Na, or Ca from Ba). On the converse side, D's have never been simulated in a highly-charged pore (people including Chung have put an ion and waters into uncharged pores). I think it is important to say that reason for the small flow in Ca and Na channels is an open issue.

Perhaps, in responding to the related question of the other reviewer, you could utter your working hypothesis (supported among others by your experience with RyR) that D scales inversely with fixed charge density. D is known to scale that way with ionic density in bulk solutions. It is plausible that in narrow pores this escalates -- salt crustals, the other extreme, are insulators despite their being made up of pure ions.

The third major point of the other reviewer is easy to address: the NE excess chemical potential is exactly what you call the screening component here. The 'excess' comes in as you move ions closer together so that they interact significantly.