There are a number of things to unpack here.
To a statistician, "significant" means "very unlikely to have happened purely by chance", i.e. we are seeing a real difference, not sampling error. To a lay person, "significant" means "big enough to matter". You are arguing that this result is not significant in the second sense.
If there are non-linearities in a system, small shifts in the mean can have a large effect. For example, a town has natural temperature range between -20C and +45C. An increase in the mean of 2C is small compared to that range. However, the number of days per year hotter than 40C might easily triple with that +2C shift in the mean (due to the shape of the high temperature tail of the distribution), and if >40C is a threshold for causing major health problems, then the small shift has a large effect.
145g might be significant in this way: a 1355g baby might have much worse survival chance than a 1500g baby. (Further complicating things, although the mean might shift by 145g, the shape of the distribution might also change. The shift could affect low weight babies more or less strongly than normal weight babies.) I don't know enough about babies to know whether that 145g shift is important or not.