Issue
This is technically not an issue, but I want to start a discussion about reaction path degeneracy for heterogeneously catalyzed reactions. When RMG discovers a reaction and applies a rate rule to estimate the kinetics, it multiplies the pre-exponential coefficient by the reaction path degeneracy. Let's assume that RMG discovers the following reaction:
^*CH2 + ^* = ^*CH + ^*H
The BEP rule has a pre-exponential coefficient of A=1e21 cm2/mol/s. RMG multiplies this factor now by 2 because of the degeneracy. When I would provide this reaction in a library, I would provide it with A=1e21 (let's assume that this is what TST gives me.) So, the rate would be lower by a factor of 2 because I do not account for degeneracy. The heterogeneous catalysis community does typically not account for it (at least, I haven't seen it). While a factor of 2 (or 3) does probably not matter much, it becomes more complex for adsorption reactions.
Assume the Surface_Adsorption_Dissociative family has a general sticking coefficient of 0.1 for unknown molecules. If this rule was used for CH4, the sticking coefficient would be s=0.1*4 or s=0.6 for C2H6, which has a bigger impact on the rate. This estimate is not correct.

To circumvent this issue, I created nodes in, e.g., Surface_Dissociation family, where I divided A by the degeneracy that is later applied by RMG. While this seemed to be smart in the beginning, it creates very messy trees. I'm currently working on cleaning up the trees, and I no longer think that this approach is feasible. It is bulky and creates too much confusion. Therefore, I want to remove the extra nodes.

Question
The question is now: How should we handle degeneracy for surface reactions? Is the usage of degeneracy correct, and has the community been doing it wrong? Should the A factors differ for, e.g.,* CH,* CH2,* CH3 dissociation? Should we treat adsorption differently compared to the surface reactions?