it's normal that you can't tilt cause you filtered all consequent nudges with tiempofiltro
if you put tiempofiltro to 2000 and nudge count interval to 1000 that means you can nudge every 2 seconds, but after one second tilt counter goes down, and till next nudge it is on 0
tilt counter makes tilt when it goes over msensitivity!
so to get tilt sometimes you can reduce tiempofiltro and/or enlarge nudge count interval
and you can reduce psensitivity and msensitivity to 1-5 not 10, putting msensitivity to 10 makes opposite of what you want to do
Before I axed the question here, I did google vpmNudge.sensitivity and found the *.vbs documentation stating:
· Property: .Sensitivity = sens
Specifies the tilt sensitivity between 0 (low) – 10 (high).
Example:
vpmNudge.Sensitivity = 5
according to the statement here, the higher the number the higher tilt sensitivity should be... hm, maybe I was wrong, I did that like 2am and at that time of day(or night) logical mistakes tend to multiply 
Will see, when I get back home from work I'll test the vbs code with tiempofiltro around 200 with count interval 1500, msensitivity and psensitivity to 1. This should increase tilt probability.
That's true for that sensitivity but msensitivity and psensitivity in core act opposite because of:
Public Property Let Sensitivity(aSens)
If pSensitivity = 0 Then mSensitivity = (10-aSens)+1 Else mSensitivity = pSensitivity
End property
It was always like that, that's not my invention - sensitivity in table and in core are inverted.