... #2 - working power of this minimum amount; I would tend to disbelieve that it lasts more than one hour and you should change for fresh developer then
One major part of the idea of stand developing - and for that matter, compensating development in general - is that very dilute developer
will be exhausted in highlight areas, while at the same time continuing to work in the shadow areas, so as to produce a compensated development - restraining the highlights, and more fully developing shadows. To put it crudely, a sort of built in burn and dodge mechanism.
It all depends upon weak developer running out of steam most strongly where highlights demand the most chemical activity. Adding new developer, or refreshing the developer in proximity to highlights by agitation will weaken this action. So, paradoxically, what you want is a weak developer that will be quite spent part way through the process, at least in highlight areas.
Adding fresh developer halfway along the process will suppress the compensating action and render the process less compensating.
The edge effects widely noted in highly dilute developer / low agitation usage also depend upon localized exhaustion of developer near highlight edges. Absent local exhaustion, the edge effects don't happen.
If one takes the agitation techniques one is first taught as a beginner, so many seconds every minute, as gospel, then compensating development with diluted and therefor weak developer and low to no agitation is heresy, but it is a useful heresy.
I might add that there is at least one other compensation strategy; two bath development, where the film is saturated with a slow acting developer in the first bath, and then, before development is anywhere near complete, transferred to a second bath consisting mainly or solely of an activator. Since the developer carried over within the film is the only developer available, it exhausts quickly in highlights, restricting development, whereas in shadow areas, development continues further, once again providing a compensated activity.
Regardless of all this, I am i bit curious about using compensated development in an intentionally seriously underexposed situation. Would you not want to build up all available densities?