[grey-area]

The problem of course being that for this to be true, housing price inflation has to be greater than general inflation

In real terms I agree house prices oscillate around zero growth (sometimes with decades long periods of higher or lower growth).

However, the leverage makes a difference. If you know bank policy is constant inflation, taking on large debt to buy an asset which tracks inflation makes sense in the long term. Over time the debt becomes less and less significant, and the asset price and yield tracks inflation/wages.

If you could borrow multiples of your wage at low interest to buy stocks for decades the same argument applies - the yield in future will outstrip the cost.