Extrapolating back to when the genome size was 1 probably doesn't make sense, but what this approach might be able to do is suggest the smallest functional geno

Extrapolating back to when the genome size was 1 probably doesn't make sense, but what this approach might be able to do is suggest the smallest functional genome size, about 10,000bp based on my reading of the graph. It's also worth pointing out that we don't, in general, have a way to know what portions of the genome are truly functional, nor even a really great definition of functional (the ENCODE consortium caught a lot of flak for theirs as overly broad). Also, almost all the genome sizes we know are from the modern day: bacteria as a class may be 3 billion years old, but they've been evolving over those 3 billion years too!