Nope, Haldanes's dilemma is a concern whenever genes move to fixation in a population. The excess reproductive capacity must exist and must be used to fix the allele.
Joe Felsenstein did not solve the problem. He merely claimed he did.
You are correct that ERV's are not observable, by defintion.
Be careful whenever you deal with things that are not observable, by definitin.
Are you saying that genetic drift cannot move an allele to fixation?
Are you saying that sexual selection does not move alleles to fixation?
Are you saying that all movement of an allele to fixation is the result of large numbers of deaths during a small number of generations?
"You are correct that ERV's are not observable, by defintion.
You seem to be having trouble with this definition. ERVs are observable. We can see sequences in a genome that we label an ERV. What is not observable is their insertion into the genome. This is because, by definition, they are passed on from one generation to the next. They effectively come from within. I thought you understood this distinction
"Be careful whenever you deal with things that are not observable, by definitin.
Nice equivocation. Unconvincing though.
Let's try a little imaginary experiment to clarify the definition.
We are in a lab, wearing some rather tattered lab coats (mine has my name across the front) dealing with some cute (according to my daughter) but smelly mice and a specific highly contagious but nonlethal virus. We know the genome of the virus and the genome of each of the mice.
We introduce the virus into the mouse population and as expected all of the mice come down with a mild flu giving us evidence that, as expected, the virus has indeed replicated itself in every mouse.
A few weeks later, all of the mice have gotten over their flu, been removed to a virus free environment and have paired off and are 'doin' their thing'.
A few weeks later a pile of baby mice are born and placed into an environment where the virus has never been and has no chance of infecting them. The genome of each mouse is checked and it is found that two of the babies have a new segment in their DNA that is identical to the DNA of the virus. We have just observed the insertion of a retrovirus into the genome of those two mice. Neither parent had that particular sequence so could not have passed it on to their offspring.
We then isolate the two baby mice with the retrovirus and wait for them to develop into adult mice. When adults we observe them 'doin' their thing' (nice that one is male and the other is female ain't it?).
Some weeks later they have a pile of babies. We check the genome of each baby mouse and find that about 1/4 of them have this same pattern at the same place. Since that sequence was passed on from the parents it is no longer called a retrovirus but an endogenous retrovirus (ERV). We have not observed the ERV entering the genome from outside the lineage, so it has 'come from within' (endogenous). We do know where it came from.
Just like your concern with selection costs, you are making the difference between a retrovirus and an ERV bigger and more important than it is.