I wondered if your poem was composed of self-validating sound packets. The Drottkvaett format requires that sound packets of six syllables contain a least one alliteration (A) and one rhyme (R). The more alliterations and/or rhymes in every six syllables, the stronger the validation.
Validation means that the sounds you hear are the sounds said by the person, who created the poem, centuries ago and thousands of miles away.
So I rearranged the poem you sent into lines of six syllables each. Then I counted alliterations and rhymes.
1. Ár var alda, þar er Ýmir bygði
2. var-a sandr né sær né svalar unnir;
3. jörð fannsk æva né upphiminn,
4. gap var Ginnunga en gras hvergi.
1) Ar var aid a par er A3,R4
2) Y mir by goi var- a A0,R4
3) San dr né sær né sva A5,R4
4) Lar Un nir jörð fan nsk A2,R4
5) æva né up phi minn, gap A0,R2
6)var Gin nun gaen gras hvergi
Rearranging the poem into six lines of six syllables, gets an exact syllabble count of 36, which is an encouraging indication that the original composer was trying to create self-validation sound packets.
Four of the lines have Alliterations.
All of the lines have Rhymes, which is another positive indicator that the composer was trying to make self-validating sound packets.
One missing alliteration may be recovered by changing “a” to “Ya, if meaning is still reasonable.”
The second missing alliteration is harder to understand. This broken sound packet may indicate that one or more syllables have been changed as the poem was recited through the centuries.
All in all, a sound set with 66% validation does imply that most of the poem has the original syllables that the ancient poet waned us to hear.
Thus, if your understanding of what the syllables mean is correct, you can proclaim with assurance that the poem is an accurate poem that the original creator of hundreds of centuries ago wanted us to hear.