There are many situations where a common 3-word term is listed in the Semantic Index as two 2-word terms. This is often the result of the 2-word term "subset" having a higher statistical value than its longer, 3-word term.
For example:
Suppose the term "business class lounge" is statistically important to your search phrase.
The Semantic Index term "class lounge" may appear instead of (or in addition to) the expected term "first class lounge".
In this case, it is possible that another 3-word term like "world class lounges" also appears in the competing search results. Therefore the two-word term "class lounge", which appears in both 3-word variants may have higher statistical value than either of the full, 3-word terms.
While it is unlikely you would use the term "class lounge" alone, the two-word instance indicates optional variability. Use whichever 3-word variant you prefer. Or both.