It's strange that this article doesn't actually mention the number of deaths for each year. They only mention the 40% increase for the younger demographic and the source says "the increase in deaths represents huge, huge numbers".
Let's assume deaths for older people haven't changed so the increase is entirely among the younger group.
x: younger
y: older
1.4x + y = 1350
x + y = 1319
0.4x = 31
x = 77.5 (younger 2020)
1.4x = 108.5 (younger 2021)
y = 1241.5 (older 2020 and 20201)
So assuming no increase in non-COVID deaths in the older group, we have 77.5 deaths in 2020 and 108.5 deaths in 2021 in the younger demographic. The actual numbers would be even lower since there have obviously been deaths in the older population.
Unless I am embarrassingly wrong on the calculation, these figures don't make sense. The source says this 40% increase figure is based on his life insurance customers who are “primarily working-age people 18 to 64”. This being just among his policy holders could explain the tiny absolute figures in the younger group. But where are the numbers for the older group coming from? His statement doesn't exclude the possibility of policies for ages < 18 and > 64. But if that's a small fraction of their customer base, it would imply a huge death rate for the older policy holders.
These numbers seem very fishy. Even if his older members are dying at astronomical rates and his numbers are correct, this is about as far from a representative sample as you could get and cannot be extrapolated to the general population. And I assume that's what the source means when he says "the increase in deaths represents huge, huge numbers."
By the way, I am not implying that there hasn't been an increase in non-COVID deaths among the young and old in these last two years. We know that's the because there is actually real data on this that the article could have included.