Hey there!
Late reply, I know... But I have been off for a while due to other priorities.
As I can see you are refering to my answer. It's been a while since I last laid my hands on WM, but I remember this case.
So, I know I might have sounded a bit greek on that reply (because I am Greek! hahaha...ha......*funny joke*

), so it's not your fault for not understanding. I will try to put it as simply as I can. Unfortunately, my pc broke a while ago, and I am typing from one that cannot handle WM. So, I will not give you a formula, instead I will try to help you understand how it works, so you make it freely yourself and in your style.

Simply put, from my emperical research on nature, my observation is that smooth sandy terrain landasses occur on the flowing parts of the landmass, or on the "insides" of the coastline, not all-around, except for some cases that the sandy landmass is high, or affected by natural phenomenons, like wind, thermal weathering, or sea weathering.
So, along a coastline:
sand ->
"insides" rocks ->
"outsides"As a start, try masking a coast erosion on the deposition output of the common erosion. And see where you get from there. Not the flow one, it won't look that good. Here is a simply (and little bad) example:
_______________ ______________
|
EROSION |-------------------|
COAST EROSION|
| flow output| |
mask input |
| wear output| |
|
deposition output|-----------------------------|
There is a better and more realistic idea, which I was working a year ago, that works with maxing an erosion over another, though simple, pretty complicated to explain... So, your best bet here is when I have the chance to show you a .tmd example...
Good luck and cheers!