Need help regarding hypothesis of branch model

Question

[rprog008]

Respected member, I am new learner of codeml and designing my experiment for analysis of 14 mammalian coding sequence based on the tutorial at https://link.springer.com/protocol/10.1007%2F978-1-4939-1438-8_4.

In my analysis, human is “foreground branch”.

I am getting confused with the hypothesis of branch model. image of the Hypothesis given in the link below

https://ibb.co/k49D9F4

My question is in first comparison (Free-ratio vs. one-ratio model), we are already testing if for a tree of s species, is ω different among lineages?

Then why are we doing Two-ratio vs. one-ratio model to test "Are the foreground branches that you specify more likely to have different ω from background branches?"

I think in both the case we are getting same result.

Please help