Burned BMS charge port @Scheppach BSG300 power station? fixable? replacements?

[Pfriemler]

Short form: At my portable power station (300Wh LiIon with 230V/USB/12V) the charge input is blown. Is this repairable? If not, which BMS could I install as a replacement?

Long form:
I've had a small powerbank (Scheppach BSG300) in use here for a few months, ~300 Wh, 230V 300W, storage cells 8p4s Li-Ion with 18650. Perfect sine. Little self-consumption. Charging is done via power supply 17,5V 4,5A (charging power according to display then about 80W) or "solar cell idle voltage 14-40V 120W" (information in the manual, display up to 121W (I have never seen more). No max input current mentioned.
I use a 160Wp module, which thanks to suboptimal orientation alignment actually never delivered more than 120W. The power station was charged like this several times last year, even in bright sunshine, without any harm.

Now in the first early pre-spring sun it suddenly smelled burnt in the room... Charging is still going on, testing with 40W. With the power supply but practically nothing more (7 watts, without display in the
display). I was quickly able to find a heat source in the BMS. Since the bank is still working fine and via USB-C-PD (is both output and input) can even be charged with 60 watts, I did not have the heart to complain about it and scrap it.
Moreover, it was quite a bargain (0.67 € / Wh), 199 Euros overall.

I have attached a schematic diagram.

The state-of-charge apparently is determined by coloumb counting, the charge path therefore is a strange way (except with USB-C-PD, where it goes directly backwards into the battery).

To the used BMS I could not find anything, there are apparently assembly variants for 300 and 500W. The power switches are 8 parallel N-Mosfet 30V 60A 2,1 mOhm (CRTM025N03L), at the charge input two of them are connected in series (why ever, maybe because of the specified 40V max input voltage?) An aluminum heat sink with thermal pad is installed above everything. This was obviously not properly tightened in the area of the charge FETs (you can see the by the non-existent indentation depth in the pad). Apparently hereby both MOSFETs were overheated gigantically once, one seems still ok, but the other one has permanently 2,2 Ohm - this explains the still enormous heat development.

A possible repair operation would be to use two identical MOSFETs in parallel on the additional on the existing additional surfaces. (The control lines of all four are parallel via one line with four single resistors.) I would leave the broken ones directly in or make them somehow make them harmless.

It remains unclear why the defect occurred. A particular overload has not taken place. The input does not have any MPPT functionality, I think a solar cell (and the AC charger, too) is simply pulled into the current limitation, so I can not imagine a charge power control function on the BMS. AC adapter is a CC/CV one with 17.5V and (measured) 4.9A down to <10V. And the input mosfets should not have any heat dissipation at all?

Whereas the AC charge input is limited to 7 watts (with the 2.2 ohms) the solar cell simply blasts more voltage into it and so it comes more voltage and so it comes to a charge, but with giant heat development and therefore useless or even dangerous.
Another option would be to directly use an alternative BMS. Unfortunately there is still some kind of current measuring shunt - two parallel R003 in in series to the battery(-), connected via two thin wires to the "mainboard". connected. In a first test the box worked without this connection.

I can't find the MOSFET to buy in a reasonable manner (I do not need 100 of them) Does anyone know of readily available alternatives?
Is there any recommendation for a replacement BMS? On A'n you can find various 4s-BMS. 30A for under 20 Euro. What makes me suspicious there is that nowhere a cooling is mentioned. But should be standard, right?

Lots of questions. THX to DeepL for translating.

Any input is welcome. Did I understand something wrong? etc.