Without any molecular evidence you could infer convergent evolution if the species being compared shared a common ancestor that lacked the trait in question.
With genomic and developmental data you would measure the degree of homologysimilarity shared between the relative factors in each species. No homologyLittle or no similarity indicates convergent evolution.
Horizontal gene transfer is predominantly found in bacteria. Although there are examples of limited gene transfer in eukaryotes, the donor is almost always bacteria. And there are currently no known cases of multi-gene transmission of complex networks.
It is interesting to contemplate how a full network might be transmitted horizontally. Horizontal Network Transfer would first require a mechanism for segregating the relevant genes from the rest of the genome. It would also require invasive machinery that allows the large amount of foreign genetic material to bypass the host's defense systems. Furthermore, integration of a foreign network is increasingly more difficult with the more genes you transmit due to positional and regulatory effects of the host genome. OnAnd on top of that, it is highly unlikely that a full complement of genes involved in a complex developmental process would function in any way resembling eye or wing development. These networks have been coevolving within massively complex genomic environments for hundreds of millions of years. Expressing a large network in a foreign cell would likely prove toxic.
A looser definition of HGT might include bioengineering, specifically stable transgene integration.