I believe they are official ones. Little metal circles pinned into the cement. Survey was done by the previous owners. They line up with the ortho imagery and not with the logical property lines (what the entire street and every property on it are aligned to).
The ortho imagery has the front of the properties too far back on one side of the street and too far forward on the other, so it is not a single property line issue. The data set [1] used has tolerances listed but it is frustrating that it translates to a survey.
The metal circles you describe are likely the official survey points and likely marked well after WGS84 World Geodectic datum | universal GPS coords .. in a neighbourhood that may have been first marked out sometime before (?)
(Perhaps the 70s | 60s | 50s ? - I'm from Australia and not overly familiar with time + architectural style from air photos in your area).
The photo image accuracy in your first link is quoted as:
* 1.35 feet at 95% confidence interval
* 3.63 ft at 95% confidence
for two seperate products .. you can read that as most displacements (of pixels (say a footpath corner)) should be less than either 1.5 or 2.5 feet (depending on product) but you should recall that some parts of a larger mosaic will have areas with larger errors and other parts will be damn near spot on.
If the "modern" GPS era monuments | pegs | nailed discs in concrete don't precisely align with the (I assume) pre GPS era road edges, foot paths, easements, etc then that can be put down to initial developer first survey doing a job that was "good enough for the 70s" (or the era of the time).
These days many land councils focus on the "total gestalt" of a series of co developed blocks - are the individual block all of the correct dimensions and areas .. (even if shifted by a foot or two from the "expected") ?
This is an extremely common issue that arose right around the globe as GPS became the global standard and many (many many) previously seperate datums and local area grids were tied together in single unified system.
Sounds like the people who built the neighborhood may have screwed up? Otherwise it's common for things like driveway replacements to accidentally (or purposefully) overextend when the real property lines are forgotten.
The ortho imagery has the front of the properties too far back on one side of the street and too far forward on the other, so it is not a single property line issue. The data set [1] used has tolerances listed but it is frustrating that it translates to a survey.
Some examples of not my house. [2]
1. https://lariac-lacounty.hub.arcgis.com/pages/lariac5-documen...
2. https://imgur.com/a/Nqkdkjj