I think the basic answer is that the sulphate concentrations are very low and therefore it takes a large amount of Ca to approach saturation with respect to gypsum. In fact, it looks like gypsum saturation would never be reached in this system. See the modified file below. I increased the amount of Ca(OH)2 reactant from 0.03 mol to 0.3 mol and added a User Graph function so that you can see the result of Gypsum SI and pH (using the PHREEQC database). I also turned off the potential precipitation of all other potential solid phases for simplicity. You can see the increase in Gypsum SI and pH with increasing Ca(OH)2 but the SI starts to level off, still below SI = -1.1. At this point the pH is well above 12. If you were to be excessive and add even more Ca(OH)2, you would find that the Gypsum SI starts to decrease, probably due to the very high ionic strength. You will also see varying dynamics in the Gypsum SI once you begin to add potential solid phases back to the solution.

SOLUTION 1
pH 3.62
unit mg/l
temp 16.0
Na 3.6
K 3.9
Ca 3.7
Mg 2.5
Cl 1
S(6) 93 as SO4
Fe(3) 3.5
Zn 0.15
Al 3.9
Mn(2) 0.61
Cu 0.23
save solution 1
END

SOLUTION 2
pH 3.44
unit mg/l
temp 16.0
Na 2.0
K 4.4
Ca 3
Mg 1.8
Cl 1.1
S(6) 63 as SO4
Fe(3) 5.3
Al 3.2
Mn(2) 0.57
Cu 0.78
Zn 0.25

MIX 1
1 0.34
2 0.66

SAVE solution 3
END

use solution 3
EQUILIBRIUM_PHASES 1
Gypsum 0 0
#Ettringite 0 0
#alunite 0 0
#gibbsite 0 0
#Jarosite-K 0 0
#Fe(OH)3(a) 0 0
#Manganite 0 0
#Schwertmanite 0 0
#Cu(OH)2 0 0
#Calcite 0 0
CO2(g) -3.64 4e-4

PHASES 1
Schwertmanite
Fe8O8(OH)4.5(SO4)1.75 20.5 H = 8Fe 3 12.5 H2O 1.75SO4-2
log_k 18
Cu(OH)2
Cu(OH)2 2H = Cu 2 2H2O
log_k 6
delta_h -56.42 kJ
Ettringite
Ca6Al2(SO4)3(OH)12:26H2O = 6Ca 2 2Al(OH)4- 3SO4-2 26H2O 4OH-
log_k -56.86
REACTION 1 Add Ca(OH)2 to the acid groundwater
Ca(OH)2 1
0.3 moles in 100 steps

USER_GRAPH
-headings mol_CaOH2 Gypsum_SI pH
-axis_scalex_axis auto
-axis_scale y-axis auto
-axis_titles "mol Ca(OH)2" "Gypsum SI"

-start
10 graph_x RXN
20 graph_y SI("Gypsum")
30 graph_SY -LA("H ")
-end
END