Using a control-register instead of a regular clock will work, there is at first no limitation for tle lowest frequency^. You should consider that the inernal capacity is very small in the area of fF (fempto Farad) and so will not be able to hold an output very long...
The PSOC device geometry input, which would be a reflection of Cgd, eg. the
hold capacitor, is 5 pF nominal. There is a droop spec in datasheet, with no value
given. You can easily test droop with simple test bed and the delsigma, acquire
a sample at full scale, and measure time it takes to get to close to ground. The
droop largely a leakage value, and you can google for MOSFET leakage
contributors, IEEE papers especially, will give you an idea if it looks like a constant
curent source dominant contributor, which I think it is.
Then use Q = C x V, I = C x dV/dT to calculate the leakge. Note it is a strong f(T).
Interestingly found another schematic including T/H and shows a different
value for hold cap, I will clear up with Cypress ( here its 4 pF) -