When a compound is partitioned between organic and aqueous phases, it becomes distributed between both phases. The distribution is not binary - it isn't
either in the organic
or the aqueous, it is present in
both (see partition coefficient), but usually there is a preference for one phase over the other.
If the organic layer was wanted, why was EtOAc necessary.
For the biphenyl you mention, it will have a strong preference for the organic layer, but some will be in the aqueous layer (remember the aq is saturated with t-amyl alcohol). The purpose of the EtOAc extraction recover the traces of biphenyl that remained in the aqueous after the first separation.
Doesn't it say to extract aqueous layer with EtOAc so we discarded the reaction mixture after adding it to hcl and the product would be moved from hcl to EtOAc.
This was a huge mistake, you threw away the majority of your product.
The prep says:
1. The reaction mixture is added to HCl(aq), then separated. Most of the product is in the organic phase from this step.
2. The aqueous phase from (1) is extracted with EtOAc.
3. The organic phases from (2)
and (1) are combined, washed with brine and purified.