I see you went with Proto-Algonquian.
------------------------
Okay, now I need a way to get Shaituhu out of Proto-Asonic. This essentially means I have to get rid of aspirated stops, add ng, s, z, and x, and get rid of w and r.
The first step is basically
Grimm's Law--aspirated stops->voiced stops, voiced stops-->voiceless stops, and voiceless stops-->voiceless fricatives.
So I have bh>b,b>p,p>f; dh>d,d>t,t>s; gh>g,g>k,k>z.
Now, I need to get rid of the labial consonants (b,p,f,m). This can be done with a chain shift that sends labials to alveolars and alveolars to velars, which isn't totally implausible. I'll justify this by saying that a thousand years ago or so, lip-piercing was common among the Proto-Shaituhu, making it hard to differentiate labials. (Now only shamans practice lip-piercing, but the language has shifted. Maybe in a thousand years labials will reappear.)
So the new rules are m>n,b>d,p>t,f>s; n>ng,d>g,t>k,s>x.
Now I need to get rid of w and r. This is relatively simple--w and r aren't really articulated clearly, so they get articulated even less clearly in the process of lenition (weakening). So I have w,r>h.
Finally, I need to get z somehow. Let's say s becomes z (i.e. it's voiced) after other voiced consonants. This accounts for dz,gz,nz,ngz. After the change, new words were coined with an isolated z, but you won't find any of those in words with Proto-Asonic roots.
Oh, and there's a glottal stop. But it's not written, and I'll assume Proto-Asonic did the same thing (most languages did).
So in 4 changes, I've derived Shaituhu:
1. Aspirated-->Voiced-->Voiceless-->Fricative.
2. Labial-->Alveolar-->Velar.
3. w,r-->h.
4. s-->z after n,ng,g,d.
I knew that Historical Linguistics course would come in handy...