Yes I have similar results with mine. The temperature is accurate when compared to other sensors I have, but the humidity is way out. When the average of a bunch of other sensors (and the reported weather service humidity) is around 45%, the Shelly H&T is reporting 69% !
In addition to the inaccuracy, you can see in the attached chart it's very slow to react to changes. This shows it placed in the bathroom right next to an aqara zigbee sensor, and the spikes are two seperate uses of the shower. This is my primary use for this sensor - turning on the fan when the humidity spikes from the shower and off when it returns to normal. You can see in the chart how slowly it both rises AND falls compared to the aqara.
I currently have an offset entered of "-24" which at some levels makes it correct, but as you can see in the chart, and as tested above by libreiot, the innacuracy is much less when closer to 100% and the offset would make it vastly underreport at this level.
I've relegated my H&T to a task that requires less precision, and am only using Zigbee sensors now (I've been really impressed by the Aqara ones). Still love my Shelly relays & dimmers, but IMHO all the battery based products suck (this is as much a weakness of WiFi than the Shelly implementation itself)