Several of the listed words don't seem to be words at all, and others are abbreviations and acronyms

I downloaded words_alpha.txt and extracted all the three-letter words, using R. Here are the first 20:

library(dplyr)
words <- read.table("words_alpha.txt", col.names = c("words"))
three <- filter(words, nchar(words) == 3)
head(three, 20)
   words
1    aaa ← reduplicated form of the name of the letter "A"
2    aah
3    aal
4    aam
5    aas
6    aba
7    abb
8    abc ← acronym
9    abd ← acronym
10   abe
11   aby
12   abl
13   abn ← not in the OED
14   abo
15   abp ← not in the OED
16   abr ← not in the OED
17   abs
18   abt ← abbreviation
19   abu
20   abv ← acronym

Problems:

1. Many of these don't seem to be words at all, or at least aren't listed in the most comprehensive dictionary of the English language, the Oxford English Dictionary, which currently lists 520,779 entries and should contain a significant portion of words from a list of merely 370,105 entries. In my sample, three from twenty words from the list aren't in the OED, which is 15%. That's quite a lot.
2. Some of the actual words are acronyms or abbreviations, which, while words, aren't what most users would expect when they want a list of words. It makes the list quite unusable for many applications.

At least issue 1 needs to be addressed.