There are quite a few stations who have misunderstood what QTH info to register in QRZ.com. Because they are located somewhere in a coastal state (e.g. Maine), they may add a IOTA group from their state to their QRZ.com info, although their QTH may be located way inland, and they never operate from an IOTA island.
When I download data from QRZ.com and I see that I have worked a new IOTA group with QTH info in QRZ.com indicating that they are nowhere near an island, I send them an e-mail and ask whether they worked from an island. In many cases they have not, and then I ask them to update their QRZ.com data because their error may create log errors for their QO partners. None of them were aware of the consequence for their counterpart logging, and all of them have corrected their QTH data on request. This happens occasionally (a times per year?) and I have not noticed any particular increase lately. Inconsistencies between grid locators in QRZ.com and locators received in WSJT-X modes are more of a daily routine. I stick to the QSO locator, which requires me to pay attention when downloding QRZ.com info to avoid QRZ.com overriding the QSO data.
I am aware of this issue because I regularly move between mainland and island QTHs. I update my IOTA and locator data in QRZ.com when I start operating from another QTH, and leave it there until I start working from the next QTH. In this way those who are interested enough in who they have worked to update their log soon after the QSO, will download correct QTH data. Those who postpone the log update may get incorrect QTH data, but that is the best I can do. I also leave a note in my QRZ.com bio page that I am/have been working from an IOTA island, so it should be possible to track my QTH.
73 Frode LA6VQ