It's far more likely that I live in a stray straw accumulation zone than straws are uniformly distributed over the entire city's gardens by some method.
[0] https://en.wikipedia.org/wiki/Faulty_generalization#Hasty_ge...
Considering that a lot of land in a city isnt your garden/house (shops, offices, parks), I would suggest that that would tend to balance out the risk of this being a SAZ.
I take your point that this is a data point of one. But if I found one straw in my garden, it is astoundingly unlikely that is the only wild straw in my city. I wouldn't start putting a figure on it from that, but it does suggest the lower bound is much higher than one.
Having never found a straw in my garden my intuition is that (not finding straws in every garden) is more likely than (finding straws in every garden) but obviously I have no valid basis to put a number of any kind on that.
Now while I think the original post is an example of Hasty Generalisation, I think the bigger issue is the generalisation combined with a sample bias. Self reported data points tend to be interesting because otherwise they wouldn't be reported. My thesis is that regularly finding a straw in your garden is interesting because it is unusual.
How do you think they generate statistics for this kind of thing? They don't count everything, they take samples and infer from that.