-ong [An(g)] : [Qn(g)] distribution

Hi.
Low Back Merger, NCVS, CVS and -or- words seem to be getting a lot attention, but what about -ONG words. 50 % of Americans and Canadians have [A] in -ONG words, 50 % have [Q] in these words.
Some people have SONG / DONG merger, some people don't.
Is there a regional map/distribution of these two pronunciations?

I guess, [A] is more frequent in CCmerged regions which have /A/ rather than /Q/ for the merged vowel (most of the West excluding some people in PNW and people with CVS + many Canadians /mostly those without the CAVS))...

But, I've heard /A/ used by many Southern speakers who don't have the Low Back Merger, so I think the distribution does not necessarily fits in the Low Back Merger dialect maps.

I guess the /A/ is the older pronunciation, the one still used in Scotland (wrong [ran(g)]) and Northern England (wrong [rAEn(g) in Newcastle, but often realized as [ran(g)])...

(Cambridge Encyclopedia of English says that /A/ in John, song, cot is the older pronunciation, rounded vowel, /Q/ present in today's RP is a recent innovation...)

Here is the list of words

along
belong
delong
dong
gong
hong-kong
lifelong
long
mong
pong
prolong
prong
sarong
song
strong
vietcong
wong
wrong
yearlong
zedong

How do you pronounce -ONG words?
Do you have the SONG DONG merger?

many thanks

ps
As for the local pronunciation, Long Island NY has [lQn(g)], Long Beach CA has [lAn(g)].

Ditto - those words would all have [O]~[Q] in Scotland and [Q] in Northern England.

In my own speech, all of those words have [Q:], which is my merged "cot-caught" vowel. (I'm from Central Massachusetts.)

<<(Cambridge Encyclopedia of English says that /A/ in John, song, cot is the older pronunciation, rounded vowel, /Q/ present in today's RP is a recent innovation...) >>

I always thought that the unrounding was an innovation, at least Wells describes it as such in "Accents of English".

From the book ''English accents and dialects (3rd ed'' written by
Arthur Hughes and Peter Trudgill:

''
Section XII. Northumberland
page 113
9. b long is /lAEn(g)/

Section XIII. Lowland Scots
page 117

5. Words like LONG, STRONG have /a/ rather than /O/ (cf. section XII), e.g. RONG /ran(g)/''
---

From the book ''The Cambridge Encyclopedia of the English Language (written by David Crystal)

page 93

''British English also came to pronounce such words as NOT with lip-rounding, whereas in the USA the earlier unrounded vowel (found as NAT in Chaucer, for example) remained.

>>Scots isn't the same as Scottish English; there may very well be a word /rang/ in Scots corresponding to English 'wrong'. Northumberland likewise has it's own traditional-dialect that may have /l{ng/ for Standard English 'long', but I was referring to the accent of Standard English spoken in Northumberland which would most likely be /lQng/.<<

The main thing though is that even though Scots and English definitively split with the end of the Middle English period, Middle Scots still was not identical to the Middle English dialects from which Southern English dialects of New English are descended from. Consequently, such could very well reflect dialect variation within Middle English, which was likely greater than dialect variation within English dialects of Late New English today.