The pool industry has historically used the Langelier Saturation Index (LSI) for plaster/gunite/grout pools to try and prevent the corrosion of such surfaces that contain calcium. I won't go into why the currently calculated LSI isn't correct nor why the Calcite Saturation Index (CSI) used in my spreadsheet and in Jason's Pool Calculator sticky is better. Instead, I want to write about a more fundamental assumption that "water that is saturated with calcium carbonate will not dissolve pool plaster/gunite/grout surfaces". I believe that this is only approximately true, but isn't the whole picture.
First of all, pool plaster isn't real plaster anyway. "Plaster" usually refers to Plaster of Paris and is calcium sulfate semihydrate. This does not appear to be what is used in pools. Instead, the underlying gunite is mostly a combination of aggregate (stones) and cement. The smoother "plaster" applied over the gunite is also closer to concrete using cement as a binder with far less aggregate. Though some of the aggregate may be limestone (calcium carbonate), the cement is mostly tricalcium silicate that gets cured as described by the following equation:
2(Ca3O•SiO4) + 7H2O --> 3CaO•2SiO2•4H2O + 3Ca(OH)2 + heat
There is also an equilibrium between the calcium oxide and the carbonates in the water via the following equation:
CaO + CO2 --> Ca2+ + CO32- + heat
where the reverse of the above reaction is how the precursor to cement is made in a kiln starting with limestone (CaCO3) -- and yes, you can see that carbon dioxide is released making cement production a significant contributor to adding this greenhouse gas to the atmosphere. [EDIT] The above reaction is very thermodynamically favorable and in fact pure calcium oxide absorbs carbon dioxide from the air and reacts in water so the above reaction is only inhibited in pool plaster by the calcium oxide being strongly bound to the silicon dioxide in a crystal structure. [END-EDIT]
So, interestingly, having the product on the right hand side be as high as possible will have the dissolving of calcium oxide be as low as possible and this product can't be higher than the solubility product of calcium carbonate or else scaling will occur. So this is one basis for using a calcite saturation index (the other being the presence of limestone, calcium carbonate, in the aggregate). However, notice that on the left hand side of the equation is dissolved carbon dioxide gas. This is higher at higher TA levels (and lower pH levels) so this implies that, all else equal, having calcium carbonate saturation would be better at lower TA levels (and higher pH levels) so as to keep the dissolved carbon dioxide lower. This means that having a higher CH and lower TA (or higher pH) might be better at maintaining cement (calcium oxide) surfaces. This is consistent with anecdotal evidence that higher CH, lower TA works better in preventing dissolving of grout (which also contains cement).
Since higher TA levels tend to have a greater tendency for the pH to rise in pools using pH neutral sources of chlorine (when accounting for chlorine usage -- so bleach, chlorinating liquid, Cal-Hypo, SWG), the above discussion is just another reason why having a low TA with higher CH to compensate makes sense. In practice, one would want a slightly corrosive (perhaps -0.2 or so) saturation index anyway so that the (30F) higher temperature in heaters and the higher pH at one of the SWG electrodes doesn't scale so readily. At a pH of 7.5, in a non-salt (500 ppm salt) pool a TA of 50 (with CYA of 30) would need a CH of 400 to be around -0.2 in saturation index. For an SWG pool (3000 ppm salt), a TA of 75 (with CYA of 80) would need a CH of 500 to be around -0.2 in saturation index. It is not a serious problem to not be at these levels -- these are merely targets.