Non-optimal positioning is already included in your "Measured (not projected) yield in Germany". Because that is the difference between a projected yield (under optimal conditions) and actual measurements of actual panels, which are practically never optimally placed.
> Then you won't consume all that it produces because you're at work during the day and there are no appliances running except the fridge maybe.
That is relevant for an economic calculation, but it is entirely irrelevant if you want to determine the point after which the panel breaks even regarding the energy used for its production vs. the energy produced by it. In that case, every single kWh counts, whether the owner of the panel economically profits from it or whether he or she just donates it to the grid without compensation.
And clearly, we are discussing the energy break-even here, as indicated by "The usual estimate to produce a PV panel is 600-1000 kWh per 1sqm."
> However from the physics/ecology perspective they make no sense at all and many of the panels installed today will never recoup the electricity used to produce them, making them a net-negative impact for the environment.
As I've demonstrated with concrete calculations, that you seemingly accept as valid as you perform the same calculations with roughly the same numbers, the EROI of solar panels even in Germany over their lifespan is clearly in the positive range. Maybe they "only" recoup 3x or 5x their investment, and not 20x, but they are a net positive regardless. Any number above 1x is.
In addition to this, as I've described in another posting here (https://news.ycombinator.com/context?id=45490555) there's the psychological side of things, where cheap and easily profitable balcony panels for everyone are a gateway drug for "normal people" to get actively involved in the field of renewable energy and must not be underestimated in their ability to open the minds of people for other, more efficient actions to get closer to a carbon-neutral energy economy. Since those activities tend to be heavily inhibited by broad refusal that's often not based on factual arguments, but simply on inertia in people's minds ("we've always done it the other way") and a certain lazyness to actively grapple with new technologies and developments, this effect is at least as important as the actual impact on the energy grid. Just like on the stock market, the raw numbers are only half of the story. Psychology is the other half.