Your updated question is still very vague, but I'm going to assume it is basically: "Why would the ancestral version of a gene be mistaken for a more recent version than the modern gene?"
If this is incorrect, please let me know and modify your question to clarify.
The simple answer to that question is that the mutations that occurred after the Ancestral gene resulted in less apparent divergence than the LCA (last common ancestor), which would cause the Ancestral gene to seemingly have more polymorphisms - and the general assumption is that the more polymorphisms (mutations) that a gene has undergone, the more recent it is.
So let's say you have the following DNA sequences:
5' - AAAT - 3' = LCA (Template)
5' - AAAG - 3' = Sample 1 : # Differences = 1 Nucleotide
5' - AACG - 3' = Sample 2 : # Differences = 2 Nucleotides
The general assumption is that the larger the difference from the LCA, the more mutations have occurred over time. So, under the general assumption, Sample 2 is probably the most recent version of the sequence.
However, and this is what I think is the answer to your question, because mutations can occur in any order and at any place in the genome, it is entirely possible for the third (from left) Nucleotide to have followed this mutation path: A -> C -> A
That would make Sample 2 appear to be more recent than it is because it essentially mutated "back" to the LCA version of the gene, despite being a linearly older version than Sample 1. In this way, an Ancestral gene can be mistaken for a more recent evolution of a gene. This is also why genomic data is never as strong as when paired with fossil records or other corroborating data that also aligns with the genomic data's proposed timeline; though the odds of a mistake being made grow exponentially less as more of the genome is compared and analyzed.
With very recent mutations; on the scale of hundreds to thousands of years, it's sometimes necessary to analyze thousands of base pairs to calculate an adequately confident answer.