At the extremes, I think it likely does have some impact on the choice to have additional children. Certainly the families I have encountered, especially those with 2e learners, seem to be slanted toward fewer children (anecdotally). I would imagine that, if parenting a child with exceptionalities in the early childhood years (already intrinsically a rather labor intensive age) also involves carting one's child around to multiple specialists, trying to tease out their needs and the appropriate therapies (among which I would include, for purposes of this discussion, appropriately servicing their exceptional strengths), parents might find the thought of "another child like this one" a bit intidimidating. And expensive. All of those therapies, even when partially covered by insurance, add up quickly.
And on integration/inclusion/mainstreaming: I'm ambivalent. On the one hand, I am a strong proponent of inclusionary practices for the left-hand tail, especially when paired with appropriate intensive services (in or out of the general education classroom) for the most critical need areas. When done well, without the flavor of tokenism that can sometimes creep in. And the long-term data on this have been that it is generally net positive with regard to providing greater access to the wider community for most of those who would once have been educated in substantially-separate settings. I have seen students at -3 SD function quite successfully on both an academic and a social level in inclusion settings. Again, when done well.
On the other hand, there are those on both extremes whose needs are so far outside of what a general education teacher has expertise (or time) to address, even with a specialized co-teacher, that differentiated instruction within the general education classroom is not going to meet them where they are or need to be. To be fair, I've had some 2e students at +3 SD also do well socially--but they've had the advantage in our setting of other points of commonality with their peers, in the non-academic side of their education, which has very different metrics for success, many of which are discrete skill-based, often both self-paced and collaborative, and reward initiative and creativity. Academically, the inclusion model hasn't been quite as effective, unless one counts the dual enrollment program, which essentially is inclusion of NT students into advanced coursework (in supported early college coursework).
A strong inclusion classroom will have both skilled general educators and skilled educational specialists (in an ideal world, for both below- and above-level learners), working collaboratively and seamlessly to bridge core content to each student at their need level. That's the theory, anyway. I've seen it work for about a 4 SD spread when two skilled teachers trust each other. I've also seen catastrophic failures. So much depends on the execution.
