It's very difficult to find objective data. Not only does COP, of an air source heat pump, depend on the inside and outside temperatures, but also the relative humidity.
I had a bit of a Google and found lots of contradictory information.
I agree the information is not that easy to find. You need to combine bits and pieces and measurements by individuals. One helpful advice that I have repeated and do it once again: differences between heatpumps (cheap vs. premium) are surprisingly small, and effect of condensing temperature (i.e., water temperature you need to get into your radiators/floor circuits/etc.) is just massive, regardless of brand. Therefore, while we roughly know the climate you live in, not knowing your water distribution temperature, we can't say whether your SCOP will be 2.0 or 5.0. Very unlikely anything outside this range. Maybe you have underfloor heating pipes and good insulation and need water temperature of +27degC. Then your SCOP can easily be 5.0. Maybe you have tiny radiators that act as the name suggest, primarily radiating the heat instead of natural convection, and maybe you need +80degC water to feel warm. Then SCOP would plummet to 2.0 or so, as you would run on resistive aux heating at COP=1, large part of the winter.
Then again, if you read a complaint that COP is bad when one has a massive hole in the wall and they are producing steaming hot water at full power 24/7 during winter, I would just ignore the whole discussion and try to find more useful data points. Unless, of course, you have a large hole in your wall, too.
I live in a very humid climate and COP does down with high relative humidity, when the air temperature is 5°C.
I can say from own experience that COP most definitely goes
up in high humidity conditions when the air temperature is +5°C. You can easily see this from the fact that Tpipe (evaporator internal temperature) as measured by the heatpump is closer to ambient temperature in such conditions (e.g.: dry air, Ta=5.0, Tpipe=-0.5, ice is forming; RH=100% humid conditions: Ta=5.0, Tpipe=3.0, it rains under the outdoor unit).
More heat can be extracted because turning water vapor into liquid releases a lot of latent heat. At +2 or so things get different - even the high humidity is not able to keep dT low enough, and freezing will start to occur. At medium humidity, freezing can occur already at +5 or so.
Defrosting is fundamentally not a bad thing: if you calculate the total enthalpy you are seeing the sum is actually
better than in completely dry lab conditions! You get extra heat by making ice from water vapor in the first place, and during defrost you don't turn it back to vapor, mostly liquid, so you gained more than lost! In reality of course significant losses are involved, e.g. because the compressor is used to transfer the heat round-trip, and because, even with fan stopped, the evaporator (now condenser during defrost) sits hot in the outdoors air, possibly in windy conditions, for a minute or two, so defrosting reduces true COP below the level of dry testing, but given decently working decision algorithms (which they usually are), the effect is much smaller than many assume.
Specifically air-to-water units are very efficient at defrosting because they have better source of stored heat, in form of hundreds of liters of warm water, than air-to-air units which can only utilize the tiny amount of thermal energy stored in the indoor unit coil (copper with aluminum fins) and therefore heat the outdoor unit with worse COP during defrost. This is easy to see from the power consumption of e.g. my heatpump, which, for the 2300W input power unit, is just between 300-600W during defrost, and it only lasts for 3-5 minutes.
But sure, there is a small dent in COP curve around +5degC. It doesn't get progressively worse, as very cold air will be dry. The loss for every defrost cycle is higher when very cold, but fewer are needed, so there is just a small jump of about -0.2..-0.3 near +5degC compared to dry testing which keeps quite constant down to however low temperatures.
A quote which I find concerning.
You can keep being a concerned citizen, but then again, this is how reality works. Your car consumes more fuel than the official numbers, just like every other car in existence, and just like they always did. Your condensing boiler performs worse compared to marketing material, I'm 100% sure about that. This is human nature. The big question is, are you making sensible decisions based on best effort of obtaining true data, or just using your concerns and uncertainties as coping mechanism?
I think those who estimated a COP of 3.5 for me were optimistic. 2.5 seems more likely, given the data I've seen.
SCOP of 2.5 in British weather, which is pretty optimal for air source heatpumping (at least from Finnish perspective), would require some massive blunder. It's not impossible to do that badly, but it would be an outlier. I'm close to 2.5 (don't have exact measurement but very good guesstimates) in Finnish climate which is significantly colder.
However, important fundamental fact to understand is that COP is not a constant over year. We talk about annual COP, SCOP, defined simply as annually produced thermal energy divided by annually consumed electricity. We do this because we are interested about how much money we are putting in the long run. Heatpumps are not suitable for those who only have £100 on their bank account and are struggling to pay a single bill on February, because of the cost nonlinearity, which makes the effect of time-of-the-year on the bill even larger than it already is with linear systems (basically everything else than heatpumps).
I've also realised I didn't take into account the fact that if I can stop using gas, which would also involve replacing my cooker with an induction hob, I can reduce my standing charge (a flat rate daily connection fee, which independent of usage), but then I have to factor in the fact it probably costs a bit more to run.
This is indeed another (albeit small) downside in gas-based systems, you rely on another contract and someone supplying you with this product every day. Compare with wood or oil (I use both as auxiliary heating methods) which can be locally stored for years, and if you mainly do heatpump, the volumes that need to be stored are modest.