Depending on meaning of "wrong". If it means that there will be facts that are not accounted for by current version of certain theory - then it's probably true for most of them. If it means that our current interpretation of certain facts is completely incorrect - such as we think X is caused by Y, but in fact X is caused by Z which has zero correlation and no causal links with Y - I would be very surprised if it turned out like this for half of the established theories. Though for those that are just being developed probably much more than half of them will be wrong in this sense. I guess it also depends on the field - I'd say it's unlikely we'd have some overwhelming change in classical mechanics, but if you talk about something like biology or neurophysiology - there are a lot of things that are not clear, so it is entirely possible that our understanding, for example, of how addiction works - would be radically different in 20 years.