Slate has a rather good article on trends regarding parenthood and marriage in modern America. It's probably not news that college educated Americans are marrying and having children later while non-college educated Americans tend to have children relatively earlier but are less likely to marry at all.
What always seems a little off to me is describing this as a new model, like capstone marriage. This pattern sounds remarkably similar to the traditional European marriage pattern to me (best link I could find in a few seconds), the experience of the last century, with early, stable near universal marriage being the exception.
To briefly summarize the European marriage the general trend was that a man would not marry until well established. For the poor this usually meant inheriting a father's farm, for the middle class achieving Master status or becoming established in trade, and for the nobility doing just about whatever they wanted.
There is a bit of a difference for women, though it has less to do with culture than it does with economic exploitation.* For the poor, the European marriage meant that families often exploited their daughter's** material production well into their 20s or even 30s. These economic activities usually involved household production, such as the clothing everyone wore, in addition to other subsistence activities. This was weighed against a woman's desire to leave her parents and set up her household. This system did also succeed in limiting early sex, but only with very high levels of supervision and opportunity for economic exploitation to make this level of supervision worthwhile. I can't imagine a modern parent being willing to dominate a daughter's life as thoroughly as a medieval patriarch did. Wealthier parents could marry off their daughter at any age, though this could be impacted by their ability to provide a dowry.
I don't see much difference between this and what is happening today. The only real difference is that since a woman has access to her own economic output she has the option of setting up a household of her own rather than having to wait to free herself from her parent's exploitation and find a marriageable husband. In cases where a woman can survive herself, but can't find a suitable mate, it is a natural continuation of this pattern for her to set up her own household independently.
If our desire is to move back to the atypical pattern of the last century I see basically two options. The more sensible one is to adopt economic policies that will increase incomes lower down the economic scale, if men can get on their own feet earlier it will make more sense for pairings to be permanent. The other option is the return of exploitative patriarchy. If parent's can retain control over a woman's production until she is married then it would be rather easy to reassert a system where both sex and marriage, not just marriage, is delayed. But without one of these two changes the real social foundation for the ideal marriage pattern simply won't exist.
I hope it is obvious that I disagree with the second option, but it needs to be said. I find the cultural explanation useless, everyone agrees on the ideal already. There is no evidence that shaming or other normative standards is effective. I often see Conservatives go on about how the evidence for liberal policies is too thin, the fact that they go on to suggest policies with absolutely no evidence or even historical basis to them just completely discredits these arguments.***
* I don't normally like the term exploitation. But economic systems where parent's have full rights to their children's production, both monetary and household, are clearly exploitive.
** Western households were far from unique in this. But most other systems had the married individuals moving in with the parental household, rather than forming distinct households. This reduced, and often eliminated, the incentives for parents to delay their children's marriage (and sex, obviously).
*** I do like cultural arguments generally, I just don't like when they are suggested as an alternative to programs which have evidence in their favor and no evidence is provided for the cultural argument. Which is how I often see them used as if cultural arguments were obviously salient for "handwave" reason. Cultural arguments require the same evidentiary standards as any other, methodology appropriate obviously; why some people think they get a free pass escapes me.