Good data gathering.
It would be good if there was some sort of survey of similar existing problems and (if any) solutions.
I would just guess that these dams are made watertight with a layer or layers of clay, and that earth fill would be put on top of that. If so, I wonder if that means that the clay layer has somehow been breached. If that is true, then what are the causes and what would the cures be? It almost seems as if a significant portion of the dam would need to be rebuilt at the top (eg above 780 feet) to fix the immediate problem. The pentultimate question would be how bad the problem actually is and if it is sufficient to monitor the problem to ensure that it does not get any worse.
In Santa Clara Valley there is Lexington Reservoir which is 195 ft. high. It has a seepage pond at the downstream toe, since the 1950’s. The seepage is enough to sustain some flow of the downstream creek through the midsummer months (at least, most of the time)n. Some seepage is normal though evidently not at such a high level as Oroville has. I have seen cross section diagrams on the net that have diagonal “seepage lines” through earth dams.
I am just an onlooker, so no expert info here.
“Lexington Reservoir”
I call it Sludgengton. Drove by it twice a weekday for 21 years. I often wondered if the inhabitants of Los Gatos and Campbell knew about the “leak”.
They design these dams with redundancies in the Zone 1, 2, 3, & Drain Zone layers, including Factor of Safety margins. However, over time, seepage can find a weakness. Thus, very careful monitoring is required in observing wet spots + monitoring internal instruments inside of the dam. Another critical monitoring function is regular precision surveys of the 100 placed monument markers on the outer surface of the dam. Thus, any differential "settlement" (compaction) would give early warning to a potential developing internal problem. If maintenance is lax, and only a few monuments are surveyed sporadically, then the dam can sneak up on you - until it's too late.
Fixing the internal cracks with "probe" injection of the sealing mix may be done if the structural integrity is sound. However in extreme cases, some dams have come to the brink of failure from severe internal core cracks. The dams were saved by (1) lowering the reservoir below the cracks (2) excavating the top core layer down to the faulting area, then replacing with a new high tech mix of clay/clayey material & Bentonite (look up Bentonite for more info).
note: for any "core" repair work, for safety, the reservoir is lowered below the elevation of concern, even for minor "probe injection" - less invasive process.
A problem such as this leakage on the dam is challenging to determine the immediate risk as a full assessment must be performed. Dams have failed exactly from these identical "symptoms" at Oroville as it may reveal a slow motion failure (i.e. an internal erosion through the core scenario)- until it suddenly escalates. The conditions in 2017 seem to be pushing this seepage/leakage as revealed by the large vegetative growth. What is unfortunate is that there have been years of notes in reports stating - this should be monitored (besides just looking at grass). Nothing has been done. Guesses are being proffered in the form of "a natural spring" - but without hard physical evidence from a full investigation.
For those new to the long discussion on the Greening Wet area (more info):