Ah always a couple different ways about it.
The "proper" way (in either case) would be to use pulldata + reversegeocode. Note: This consumes 0.04 credits for each reversegeocode statement that is run, so consider your survey traffic and credit limitations. This post has a good example for how to set up a single reversegeocode and use pulldata(@json,'..') for separate fields. See the attachment in the initial question - Survey123: Reverse Geocoding. At its most basic setup, it might look something like this in the calculation field to pull the city (assuming you're using the standard geocoder):
pulldata("@geopoint",${yourgeopointfieldname},"reversegeocode.address.City")
You could embed this within an if statement, or you could have a hidden calculation field if it helps you understand the process step-wise.
The alternate method I might suggest would be a compound if statement. In this case, it checks if the address is within Zlatar. (In other words, if address contains Zlatar AND it contains Bistrica, then it is not in Zlatar.)
if(contains(${address},'Zlatar') and contains(${address},'Bistrica'), 'no', 'yes')