That's a great view, I think most of the economy schools are taking this view, academically it is quite nice
I have a different view
When people sell product and get money, they save the money, but they did not save the value (the product was consumed or degraded), money was hold as a "proof of work"
It's easy to see: Saving do not increase the society's consumable resource, it just increased society's "proof of work"
So, in the example of the $5 million house, A paid B $5 million to buy the house, there are $5 million worth of "proof of work" transferred from A to B, but the real wealth is only that house, those $5 million worth of "proof of work" do not correspond to enough tangible wealth in the society
Same house could worth 5 million one year and 3 million another year, but the house do not change, only the exchange value of the house changed
The saving happened when he made the product but didn't consume it himself, not when he sold it.