Measuring house prices through time is not easy. You can nerd out and read how the repeat-sales-index was designed to get a handle on some of the problems it seeks to fix. We built it in Australia but found the very long hold-periods for houses and renovations between sales held it back.
Other methods have been designed which work well in different situations. However no method comes without constraints. One of the bigger challenges now in many markets is when houses in the top suburbs or top streets are held back from listing. What I can see happening now is many areas and suburbs with a much higher proportion of listings below the median. This will drag the prices down but not necessarily the values. Compositional bias will create a bit of havoc for us in the months to follow.
One interesting outcome will be how prices drop and then bounce back quickly in line with higher ranked streets or higher ranked suburbs re-entering the market.