Testing a theory based on the that dslr thread that it's not forwarding to vm correctly because a number equaling your gv # is getting passed as the caller id which somehow causes issues.
Theory is to pass on a different caller id when placing the call, which, according to the dslr thread, in asterisk fixed the issue.
So, you would update
{911:sp3},{933:sp3},{([1-9]x?*(Mpli)):pp},{(<##:>):li},{(<#:>):ph2},{(<**70:>(Mli)):li},{(<**82:>(Mbt2)):bt2},{(<**81:>(Mbt)):bt},{(<**8:>(Mbt)):bt},{**0:aa},{***:aa2},{(<**1:>(Msp1)):sp1},{(<**2:>(Msp2)):sp2},{(<**3:>(Msp3)):sp3},{(<**4:>(Msp4)):sp4},{(<**9:>(Mpp)):pp},{(Mpli):pli}
with
{911:sp3},{933:sp3},{([1-9]x?*(Mpli)):pp},{(<##:>):li},{(<#:>):ph2},{(<**70:>(Mli)):li},{(<**82:>(Mbt2)):bt2},{(<**81:>(Mbt)):bt},{(<**8:>(Mbt)):bt},{**0:aa},{***:aa2},{(<**1:>(Msp1)):sp1(12341234567>$2)},{(<**2:>(Msp2)):sp2},{(<**3:>(Msp3)):sp3},{(<**4:>(Msp4)):sp4},{(<**9:>(Mpp)):pp},{(Mpli):pli}
When placing your test call, dial **1{number}, where number is one of those numbers that created issues for you in the past when unanswered.