OK, my bad. I will accept the correct belief system.
BMS's must have large amp mosfets to be able to shut off the current, because loads and MPPTs should not have simple on/off switches that can be trivially controlled by a very low power relay.
Chargers should control the charge because a BMS has a big mosfet to snap off the current, because the charge controller knows nothing about the individual cells. They should also be able to hold a voltage and taper the current because doing on/off and letting the sun taper the current is much too complicated. The battery would get a headache with that juice flowing +, then -, then + again at a frequency similar to what my BEV battery experiences every time I drive it. It's bad enough that we have clouds, and day/night, doing this cycling. Plus it is ideal that these controllers that know nothing about the SOC determine when to stop because it is most intelligent to stop at 100% all the time. It would be a waste of capacity, not to mention all that intelligence, to charge to say 90% every day and do that tapering stuff only once a week or so. This is why they call them "intelligent" controllers, obviously.
It is imperative that we always buy an MPPT because a lot more circuits, with more software, and a higher price, is always better than less, less, and less. Well, 99.9% of the time, which is good enough for all of us.
Also it is important that we ignore heat sinks. Whenever we are trying to eek out the most efficient system, we should not question electronic components that produce heat. After all, heat does not affect circuit longevity or efficiency.
Hopefully this will keep the correct belief system police from knocking at my door and locking me up. I appreciate all the help from those trying to get me to comprehend the current regime.
I apologize for being so thick.
A BMS needs either mosfets OR contactors if you are concerned about mosfets failing. IMHO, one of those two is a MUST.
A BMS can control charging devices, such as an MPPT controller or (usually more importantly) an alternator regulator. But the BMS is NOT controlling the charge parameters, just turning them on/off. The MPPT is still it's own intelligent device, and should still determine when full charge is, and stop charging accordingly on its own. The MPPT still goes through the bulk and absorption phases. And the BMS is still the emergency last line of defense. IMO, a relay is still needed, because there are examples of MPPT controllers failing as a short circuit to the solar panels, and no "stop charging" command is able to stop them. Or, an owner/installer programs the MPPT incorrectly. And the relay/fets also protects from over discharge, which control of the MPPT offers no protection from.
Plenty of people use PWM controllers, and a few use basic on/off controllers. They work. The MPPT gets more power from the panels, that's all.
The sun tapering off everyday lowers the voltage, so unless you have reached absorption long before the sun goes down, voltage will drop below the cell voltage and stop charging early. This doesn't harm the cell at all, but doesn't result in a full charge.
Ohm's law proves your statement incorrect, and that you fundamentally misunderstand the article and MPPT controllers.
V=IR
Voltage from a panel will remain constant when open circuit, regardless of irradiance. In that case R=infinity, so V will be the maximum voltage the panels can produce. When you apply a load, ohms law applies. If the load is constant (for example if the panel is connected *directly* to a battery) R is not changing and you cannot reduce current without voltage also falling. That is electronics 101 and basic physics.
What you are missing and the article is discussing is that there is a power curve, where if you adjust the load resistance, both voltage and current will change. As you adjust R, the *Watts* out of the panel changes, even if irradiance does not. Now imaging constantly changing R such that you are always getting the most *power* from the panels. This is the case where voltage doesn't change much, where you are constantly changing R to get maximum power from the panels. This is also what an MPPT controller does (maximum power point tracking), it constantly changes R such that the panels are always producing the most power possible. And THAT is why an MPPT controller produces more power than either a PWM or a direct connection, even given the loss from heat.
If you put 2 100 watt panels next to each other, one connected to an MPPT and battery, and the other directly to the battery, this is easy to observe. You could then measure the panel voltages, and the panel connected to the MPPT would be probably around 18V or so. The panel connected to the battery would be close to the battery voltage, maybe 13V. And more power would be flowing into the battery connected to the MPPT.