Children of Who, Exactly?

Commentator Anonymous raises some fruitful questions in the comment section for
http://chariotofreaction.blogspot.com/2012/04/dungeons-and-dysgenics.html

I've been considering a question lately of great personal importance. Perhaps I'd like to frame it first.
Lots of people here would probably consider genetically enhancing their kids once the technology is at the level necessary. I've been wondering lately what "your kid" means. It's pretty straightfoward in a natural birth. But if you start changing the genes, to what extent does the fact that the base material was yours still make the kid yours?
What is the difference between a substantially genetically altered child from your sperm and a child from another man's sperm?

As an aside, it is probably a good idea to pick a pseudonym for such communications.  It doesn't even have to be unique, I have several that I use in non-overlapping domains.
I made a first pass at an attempt to address our esteemed commenter's questions, but I think to do it justice, it will require a post, perhaps more than one.

Let's start by discussing in a cursory fashion the genetic enhancement technologies that are likely to show their heads in the decades to come.  There are two big narratives out there insofar as how people inherit capability X.  The first is typified by the article below---specific genetic variants contribute pluses and minuses to said attributes, possibly with some weird nonlinear or multiplicative effects.  This is the big meta narrative behind the human genome project as it applies to what we term genetic 'enhancement'.
http://www.futurepundit.com/archives/008589.html
(Article describes several single gene variants that are associated with larger/smaller brains and higher/lower IQ)
The second narrative comes to us by way of Cochrane, although it has a long pedigree.
http://westhunt.wordpress.com/2012/04/10/more-thoughts-on-genetic-load/
This narrative is basically that it isn't so much the variants of the genes that do the damage but rather the accumulation of mutational load.  Applied to intelligence, it would imply that a 'spell checking' protocol---probably akin to the defect detection process used in semiconductor manufacturing---would result in very substantial improvements to that capability.  People with higher intelligence, in this narrative, mostly just have genes in that area that are just less broken by mutational vandalism.
This, by the way, is a REALLY old narrative---most ancient cultures believed that they were the inferior descendants of much more awesome ancestors, who typically had much longer lifespans and were more capable in a host of ways.

The good news is that it is likely that we'll have a much more clear picture of the extent to which each of these narratives is true fairly soon now that sequencing has found its own analog of Moore's Law.

If the answer is mostly the first narrative, we'd expect to see productized implementations of Preimplantation Genetic Diagnosis deployed, probably in the 2020-2030 timeframe.  They'd probably evolve into some sort of best 1 of N protocol, where the prospective sperm and eggs were filtered according to the best information available.  Would a child born of this protocol be still yours and your spouse's?  I'm inclined to think so.  You'd expect to see this a lot for children who are already born using IVF and other fertility treatments, making them likely the beta and alpha testers of said protocols.  It's also exceedingly unlikely that such technology can be stopped politically since there's no bright line to form a useful Schelling fence behind.  You can talk about remediation (good) versus enhancement (bad), but the line between them is very blurry indeed and there's every incentive to blur it beyond recognition.  Plus it is unstable from an engineering standpoint, every enhanced child creates pressures for more to enhance theirs as well.  Such a process is already used if memory serves for Tay-Sachs, so it's unlikely that the genie could be crammed back into the bottle even should we wish it.
Philosophically, on this, say you selected one combination of sperm and egg from, say, 25 possibilities.  All of them came from you and your wife, so the child born of it is still obviously yours.  You've loaded the dice, but God still casts them.  I bet you loaded the hell out of the dice when you selected your mate too, you scoundrel, I know I did.
How much 'better' on average would such children be?  Honestly, I don't know, but does it really matter?  If you were to exceed the Duggars and went ahead and actually had all 25 potential children, and one was clearly 'superior' insofar as you or society deems to define it would there be any innate evil in this?  I don't think so, and I think most parents with large families understand that some of their children 'rolled better than others' even though the proverbial dice were the same.  None, in as much as I know, have called for the Handicapper General.

The second narrative is far stickier in its implications if it is the dominant one, frankly, even if it has equal weight to the first in practice.  I believe I'll give it its own post.  As I alluded to above, it has some distinctly Antediluvian overtones to it.  It raises serious questions, not just of when a child ceases to be 'yours' but where does the boundary of 'You' versus 'Not You' lie.  I'll wrestle with that in another post, when I am less fatigued.