I haven't been awake enough to answer those questions, but I'll get to them sometime soon. In the meantime, here's a thing on the script. It starts by describing the Kett script because that's where the Arve script comes from.
Some information on the Kett/Arve script. I start off only covering Kett, because Arve gets very complicated, as is to be expected.
The script is vaguely a syllabary, but there are only two vowels, transliterated <a e>, and one null vowel column. All other vowels/diphthongs (depending on how you want to analyze them) are written by adding an extra syllable containing a semivowel. So, for example, the Kett word keréu 'wait' is written ke-re-w. The rules are different for unstressed vowels, but I'll get to that later.
Consonant clusters are also written with the null-vowel column: céck 'rule' is written ce-c-k.
The null vowel column, however, is a relatively recent innovation. An older method of writing consonant clusters and vowels besides /a e/ is with an echo vowel and a vowel cancellation diacritic, transliterated by capitalizing the vowel canceled. The two previous examples would thus be written ke-re-wE and ce-cE-kE. This method is still used to a certain degree in Arve.
There is also a rounding diacritic, transliterated <W>. This is only used on the consonant columns /t k j/, to form the consonants /t͡ʙ̥ pʰ ɥ/. The words varétw 'good', pes 'but', and léyka 'think' are written va-re-tW, keW-s, and le-jW-ka.
The gemination diacritic, transliterated <:>, indicates that the next consonant is geminated.
The last diacritic, the consonant canceler, romanized C, only appears on word-initial characters, and cancels an inherent /h/. For example, the words ḩéng and abhéit are written He-ŋ and HaC-b-he-j-t. Native vocabulary contains only words with initial consonants, as does the native lexicon of Arve; all words with initial vowels come from Kannow or unknown languages. (However, Kannow later appended initial h- to vowel-initial consonants, presumably under influence from Kett.)
For a longer example: (this text is very old and probably not even close to grammatical anymore, but whatever)
Ketwlá han cékcan máru.
ke-tW-la ha-n ce-k-ca-n ma-r-w (note: Kett has two vowels in stressed position, but five in unstressed position. The other five correspond to eC diphthongs / vowel+semivowel clusters. They are written as consonants.)
Ang láncapean ullá ammárung.
HaŋC la-n-ca-keW-ḩa-n w:-la HaC:-ma-r-w-ŋ
Láncakkoman cécak ve mebérra.
la-n-ca:-k-m-H-Ha-n ce-ca-k ve me-pe:-ra (note: <o> = next sonorant is syllabic. this is written by _ H)
And now for Arve. (Well, Proto-Arve.) Here's the fun part.
The system has to be stretched to write nine vowels (æ ɑ e ø o i y ɯ u) with two lengths (short, long -- overlong vowels developed entirely from compensatory lengthening after simplification of consonant clusters, so they're not relevant here), and six diphthongs (ei øy ou ie yø uo). Also, I should point out that <ü y> in the romanization are reversed from what you'd expect: <y> is /y/, and <ü> is /ɯ/. But the two vowels merged, so it shouldn't be much trouble.
æ - Ce DA
ɑ - Ca
e - Ce
ø - Ce DAW
o - Ca DAW
i - Ce DE
y - Ce DEW
ɯ - Ca DE
u - Ca DEW
The C slot is filled by the consonant directly before the vowel, and the D slot is filled by the consonant directly after it, if there is not a vowel directly after that. If there is, D is filled by H and the vowel gets a C diacritic.
For example, the word vöntus 'alcohol' is written ve-nAW-ta-sEW, and the word rätt 'place' is written re:-tA.
Length: The short length is written with the gemination diacritic. Steck 'wheat' is written s-te:-k. Long is default.
ei - Ca jA
øy - Ca jAW
ou - Ca wAW
ie - Ce jA
yø - Ce jEW
uo - Ce wEW
For example, the word huon 'hand' is written he-wEW-n, and the word zeis 'hope' is written ca-jA-s.
Eventually, the null diacritic will fuse with the characters and form a few new sets, although the modifications will be mostly predictable. I might do the same for W+null diacritic clusters, but probably not.
Consonants: The inventory is /p t ts k b d g v s x χ m n l r j w/. The voiceless stops, except /ts/, are written with a modified form of the gemination diacritic, transliterated <S>, but the rest are written with the characters for p t c k v s H h m n l r j w. There's also a glottal stop, but it's written with a diacritic transliterated <?> and applied to H_ characters.
For a longer example:
En enz sjö-tjargut, kes enter lauch.
He-n He-n-c s-je-HAW?-tS-ja-r-ka-tEWS, keS-s He-n-teS-r la-w-H
Kes treir räud skand ast.
keS-s tS-ra-jA-r re-wA-t s-kaS-n-t Ha-s-t
Now I just need to develop the actual characters.
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.