Rather than trying to find fancier ways to define the lines, are there lines at all? Perhaps fuzzy grouping opens up more powerful models. What began as co-evolutionary theory is way more continuous.
I was at SFI in the 90s, before the big push mentioned in this article. I remember a conversation about it, discussing perhaps there isn't such a thing as an individual. It wasn't even said as if radical (e.g., by that point the Selfish Gene (1976) and The Extended Phenotype (1982) were well established, and way behind any 'cutting edge'). So I suspect it isn't quite 'finding the individual' that the article author keeps focussing on. But more 'models that abandon individuality'.
I generally suspect that a problem with large swathes of human thought are discretising continuous reality. That's definitely a core SFI view too. Or at least was in the 90s.