I don't think that I can take the analysis of that second example any further. So here goes :-
Atomos in a 365 1/4- day year (i.e. 197,760,960)
A B C D E F G H I J
decies novies centies centum millia et quatuor mille millia et DCCCCXXXI LX
((A + B) * C * D * E) + (F * G * H) + I + J
= ((10 + 9) * 100 * 100 * 1000) + (4 * 1000 * 1000) + ((500 +400 + 30 + 1) * 1000) + 60
= 190,000,000 + 4,000,000 + 931,000 + 60
= 194,931,060
Byrtferth's equivalent is :-
D E C A E B C E {F?} G H
an hund þusend and hunfteontig siðon tyn þusend and nigon siðon hundteontig þusend and þusend siðon þusend
I J
and nigon hund þusenda and an and þrittig þusend and syxtig
NOTES
a) The Latin number is inaccurate, representing 194,931,060 (not 197,760,960), which is approx., but not exactly, 360 days
b) Byrtferth's number appears to be an attempted and misunderstood translation rather than an "absolute number"
c) It misses out the "4" of four million, that appears in the Latin
d) It is not necessarily a correct translation, even taking (c) into account (I think that he misunderstood the beginning)
The Latin number-suffix "-ies" represents "{number}times ... " , However, because the first number "10" is followed by a smaller number "9", the pair together actually means "19 times". The alternative option being better represented by "novagiens" (ninety times).
The Latin expression therefore begins "19 times 100 times", but Byrtferth seems to have it as "10 times the later expression plus 9 times the later expression plus 100 times ... etc"
For the Latin I used a transcript of "Liber de Computo" Book 10; Rabani Mauri [AD820], who I think was one possible source for Byrtferth
Now! It's time to repair the brain cells with a quick pint!