Saskia

The reasoning you are after is about structure, and once you have it the contradictory rules stop contradicting.

What you see as fuzz is the fruiting part. Underneath it the mould has a root network — hyphae — growing through the food, and it goes much further than the visible patch. How much further depends entirely on how dense the food is.

• Soft and high moisture — tomatoes, berries, peaches, bread, soft cheese, cooked leftovers. The hyphae spread easily and fast, so by the time you see the patch the network is well beyond it. There is no safe margin to cut, because you cannot see where it ends.
• Hard and dense — firm cheese, carrots, cabbage, hard salami. The structure physically resists the spread, so the network stays close to the visible growth. This is where the "cut around it" rule comes from, and the usual figure is about 2-3cm on all sides, with a clean knife that does not touch the mouldy part and then the good part.

Your cherry tomatoes are firmly in the first group as individual fruits. But that is about the mouldy one, not its neighbours.

"You were expecting skillet behaviour" is the whole thing, isn't it. I have been trying to make it do the job the flat one already does well.

One trap worth knowing before it catches you: a hole that takes an M5 bolt is not drilled at 5mm. It is drilled a bit under and then tapped, or it is a moulded thread.

This matters because people measure the hole, get 4.2mm, and go looking for an M4.2. There is no such thing. If you are replacing a screw, measure the screw. If the screw is gone, take the thing to the shop and try bolts until one turns freely by hand.

Your suspicion is correct and the honest framing is: the phone wins on processing, the camera wins on physics, and which matters depends entirely on the light.

The phone's advantage is that it takes a burst of frames and merges them, then applies tone mapping tuned by people who looked at millions of photos. That is why the sky is not blown out and the shadows are open. Your camera took one frame and handed you exactly what the sensor saw. Comparing a finished image to raw material is not a fair fight, and doing the same processing to the camera file closes most of that gap.

What the phone cannot do is collect more light. The sensor is small and the lens is tiny, and no amount of merging changes how much light arrived. In good light there is plenty and the difference is invisible. As light drops, the merging has less to work with and the gap opens fast.

So: bright daylight, the phone often wins outright. Dim indoor light, it is not close.