It can definitely be considered for anything built prior to late 2000's, where the re-investment question already looms and the performance is clearly low. Likely cost-prohibitive for a full replacement on anything there onwards.
New feasibility studies, community consultation, disposal, tear-down costs (which are hefty, around $35m for a FCWF, for example), funding costs, EIA's, transportation, business resources/wages, grid connection costs, downtime, new towers and of course the new turbines themselves. That's hundreds of millions for a reasonable uplift into a soft energy market.
We build these to last, and it's important for the overall health of the industry (our all-important social capital) we don't just tear these down every decade.
A refurbishment is definitely an option for aging wind farms. It's a kinder sell.