You have a hard time sorting the wheat from the chaff when bankers set out to praise themselves. On one hand it is evident that the ability to finance your home Is of value – otherwise nobody would own their home – not until they were well past the time when they needed it. The alternative: Pension funds owning rental property has been tried.
One of the results of WW2 was that housing in Germany was severely damaged: A whole generation’s savings was reduced to rubble. To this day the typical German worker and white collar employee rents his dwellings. There is no social stigma associated with paying rent as opposed to servicing a mortgage. The downside to public housing (or pension fund investment property) is rent control, obnoxious neighbors and limitations to how you want to actually live (you have to love the wallpaper issued by the landlord).
The one objective data point you have is when it is viciously obvious that the financing doesn’t work and the housing market collapses. The advantages to the debtor of a functioning real estate financing are priced into the property value. Building prices are no guide, as the building cost is as much a function of property value as the other way round.
So the estimate comes from the price difference at the breaking point and the price at a reasonable stable price level with a functioning housing market and real estate financing.

Denmark HAD in the mid 1990’ies a well functioning housing market and a world class system for financing the purchase of a home. That is probably more than can be said for the USA. Not that the American housing market didn’t function with stability.
To measure the relative merits of the “good old days” way of muddling along you need to index home prices relative to the breaking point: I.e. the breaking point is set as index = 100.

On the graph you can see that both in the USA and Denmark there was stability in the mid 1990’ies. The distance to the breaking point was 3-3½ times the stable level. The most precise figure is probably 3 times stable level, as the higher figure of condominiums is highly influenced by tax speculation by wealthy parents waiting to cash in on raising property values by fronting their offspring on the housing market.
The comparable figure for the United States is 2½ times stable price level to breaking point. This leads to the conclusion that the difference in measured in property prices (3/2.5=1.2) is about 20%. That is a middling finance system like the American cost the American homeowners about 20% of property value. Presumeably the investors will get a similar 20% less yield on their investment during normal times.
In times of crisis – well – what is the price of chaos? Chaotic?