Most Investigate Ingress posts have 3 codes hidden in them and it isn’t uncommon for single codes to have multiple passcodes hidden inside. [2] is too short to easily have 2 passcodes inside and this code has more than enough space to hide an extra passcode. Then all we need to know is which information is used for which code. Because the numbers are repeated there is a strong possibility that each group of numbers is part of the same code, so to start try only splits that keep those groups together. The first obvious one would just be numbers and letters.
[1a] 666 8 66 55 2 22 8 22 8 55 8 666 5 111 5 66 2 22 8 6
[1b] twrhaforcrsngniafoo
We will start with [1b].
This is what gives us a good indication that our split was correct. Converting letter pairs to numbers make something very close to a code, only the keyword appears to be messed up.
2rha4rcrsng9a4o
Google is giving us nothing good so maybe we are missing something. Taking another look at the keyword rcrsn it only has common consonants. What if all the vowels were just taken out of the keyword? Looking through a keyword list there is a keyword that fits.
2rha4recursiong9a4o
Now for [1a].
At first the [1a] code looks like a phone cipher but 111 messes that up. The numbers are always limited to triples so maybe the number of repeats minus 1 could be added before the number to get numbers 0-25 or 1-26, no luck there either. How about keyboard, number of repeats it the row and the number itself is the column. Bingo.
nihgwsisigintzthwsiy
9hgw6sigintz3w6y

Ignore that there are too many numbers for standard code format and organize the code into 5 rows.
4mw
cs9
j4e
q09
38b
Reading down from top left we get:
4cjq3ms408w9e9b
Which is in proper code format with an invalid keyword. With a quick google search of ms408 we are lead to something called the Voynich Manuscript, Voynich is a known keyword.
4cjq3voynichw9e9b