I too have been pretty busy / distracted, and have only done limited testing lately (I try when I'm traveling 15-20+ miles) and if it's a new area, I set my main phone on roam (i want em all
) to get a feel for how many are there.
so here's a theory dump
I'm not sure if it's the voice that's getting recognized though.. I'm suspecting a difference in 1x & evdo (i get more sprint towers in 1x) and who knows, that might be the "culprit"
so here's rough results for a given 20-30 mile stretch I've tried @ different settings just in the regular phone settings area.
1 = sprint/automatic with hybrid network
3 = roam only with hybrid network
5 = sprint/automatic with 1x only
like I said I've been busy, and some of this was done on the fly (no notes) so something else may have been different (route, setting, ect...) also I drive alone, and I try to still watch the road.
so these are "safe" setting changes (roaming charges may apply) that likely "fix" the issue. the more testing "we" can do the better it can be nailed down..
("I personally still want to figure out exactly what in "field trial" changes to collect a cell)
EDIT: I was using advanced config to end my data sessions after 1-5 min (again inconsistent)