Sounds like the meter definitely took a dump and that certainly explains that end of it.
Battery voltage "normal" under no load but unable to supply current is symptomatic of a damaged cell in the battery somewhere if all of the connections and wiring check out. More than likely what you'll find is that under load, one will still be trying to hold it's end up, while the other one goes silly, possibly even completely "mirroring" the good one with an inverse voltage across it's terminals.
I wouldn't use the electronics present to try and test for these conditions because the unpredictability could cause damage to the electronics themselves.
Rather, disconnect everything from both batteries and give them a simple moderate load such as a headlight across each one independently and see if it can hold it up, and what the terminal voltage is while it's doing so. The one with the issue should reveal itself by not providing enough power (or any at all from the sound of it) to light the headlight up, and the meter will confirm the degree of the sag.
On to the cause from how I perceive it...
I think they serve a rough life as 75% down is pretty significant for lead acid, and while there are many factors that contribute to cell damage in batteries, they're not all "equal". You're in a portable environment there for one. Discharged cells are more fragile and while AGM is physically more rigid in construction, they are still susceptible to vibration and shock... The lower the SoC, the more pronounced this is. There's a good chance there's a crack in the internal electrical structure somewhere in a cell, which because they're all in series, the limitations of the weakest cell affects the entire battery.
I screwed up somehow with the math on the theoretical "2C charging rate"... Crossed the streams... Lol I was tired when I was writing that... But you're still looking at C/5... and this is still another strong potential factor because it exacerbates even minor differences in cells (unavoidable), which leads to runaway imbalance. Weaker cells will reach both charged and discharged states sooner than the stronger ones, which ages them quicker yet because they spend more time at both ends. Sealed cells are particularly vulnerable to overcharge as excess gassing causes pressure to build, and even tho some have catalytic recombination devices in them (to turn hydrogen and oxygen back into water), these can easily be overwhelmed by excessive rate... And unlike in a flooded cell, once it's lost, the water is gone forever.
Extreme rates also can cause warping and buckling of plates due to expansion from hot spots that form... Not only leading to cracking but even internal shorts... I probably don't need to explain where that can lead...
However since you asked, yes an "event" would refer to, worst case, catastrophic failure of a cell... Explosion, etc... Best case, internal damage resulting in a shorted cell that essentially is just "bypassing" itself... And anything in between. In any case, the early demise of the cell in question, and by extension, the battery it's part of.
Charge rates need to be limited to allow the chemical reactions in the cells to be able to keep up with the current flow causing them to happen. This is what the "absorb" stage is about in charge controllers. The current is allowed to taper off while the voltage is held steady because as the cells reach full, the number of places these reactions can occur become less and less, and the places where they've taken place in full, the energy gets expended splitting water, making heat, and causing other unwanted reactions that slowly but permanently degrade [namely positive] plate material.
Typically, lead acid is designed for C/10 charging rates, and while some excursion above this can be tolerated, and in some cases even designed in for a specific battery's purpose, higher than this is generally not a good thing.
So while you're not looking at 2C (which would be *20 times* this "standard rate"!), C/5 is still very high to be using regularly. If the controller has means to limit (it may not, being only PWM), you may want to consider dialing it back some to limit those peaks. Unfortunately, this is of course a double sided coin... The controller doesn't know the difference between the battery and an ongoing external load... So its a balance. With the whole "2C" thing behind us, which I again apologize for... Sigh... You may be able to just "ignore" that for your use case. Just keep it in mind for general purposes.
Discharge rates are a little different because the source of the energy is the battery itself and there's a degree of self limiting that occurs, but it still can be problematic. Depending on the physical construction of the plates and cells, the rate is limited by some intentional combination of the plates and the electrolyte, with the battery's intended purpose in mind. Also, during discharge, gassing is *much* less a factor and so more of the plate area stays in contact with the electrolyte, helping reduce localized hot spots by way of improved cooling. The reaction during discharge is also endothermic (you'll never notice the cooling effect however because ohmic heating overtakes it).
In a battery that's meant to crank an engine, capacity limitation is imposed by the plates, with an "excess of acid" (relative to plate mass), as they're designed to deliver a lot of current for a short period and then be immediately recharged.
Deep cycle has much more plate than acid relative to a cranking battery but with a lower instantaneous surface area in contact with the electrolyte, since peak current delivery isn't as important as the ability to deliver power over long periods of time. This reduction in active area increases the internal resistance of the cells however, so when an attempt to draw large amounts of current from a deep cycle cell takes place (relative to the cell capacity), it generates more heat, which just as in charging, can result in plate warping and other identical threats that too high a rate can cause there.
The so called "Marine/RV" batteries are a hybrid between these two extremes, and for anything outside of small light loads (again relative to the capacity) for comparatively short periods, and starting smaller engines (outboards, generators etc), they're generally considered "not great at either one". They can't deliver quite the punch that a "true" cranking battery can for rolling an engine over in the cold, and they don't stand up to the deeper cycling as well as a true deep cycle can either. This doesn't mean they don't have their uses, they do, but it's a bit more limited than it would appear, and so they endure a little more abuse overall by being selected for jobs they're not particularly suited for. The price tag is usually the alluring factor...
Hope this helps some... I know it was a little drawn out, but figured I'd go ahead and cover all the bases on this one. These "little black boxes" we all rely on are a lot more complicated than they'd appear on the surface, unfortunately.
One other thing comes to mind with the advent of the pics... That's a pretty simple and clean layout... But other than the jumper, the wire sizes there look a little anemic for the loads you're saying are in use... May want to look into beefing them up a little... To reduce losses even if they're holding up safety wise.
Steve