I may have found a way to know the actual voltage that the Growatt will resume charging after it has stopped charging. However, this is only for the ultra adventurous types as it requires you to read the internal Growatt registers via the MODBUS protocol/interface. The document that describes this can be found at
https://www.photovoltaicsolar.in/Growatt_Manual/MAX Series Modbus RTU Protocol.pdf. If you look under Register 1007 in that pdf file you will see something called Vbat start for charge. The description says "should charge when lower than this voltage". The units are in 0.01V with a default of 5300 (probably 53V).
While I have not yet tried reading the Growatt registers using MODBUS, I have been totally successful doing this exact same thing for the Midnite Classic and reading its internal registers using an existing Arduino library. I seem to remember seeing an Arduino library for reading the Growatt registers using MODBUS, but to not remember where I saw it. If you are interested, you will probably be able to find this Arduino library or something similar by searching the web for something like "Growatt modbus registers" or something similar.
If you don't want to go to all that trouble, you may want to assume that the default of 53V in the Growatt MODBUS pdf is the actual voltage threshold the Growatt uses to restart the charging cycle. If you make this assumption, then this could/should impact your Float Voltage value in the Growatt Setting #20. The default Float voltage for Setting #20 is 54V as stated in the Growatt SPF 5000 ES manual. So with all these assumptions, as long as you set your Float voltage at or above 53V then the Growatt should trigger a new charge cycle when the voltage drops to 53V.
The main problem with assuming that the 53V is the actual trigger for starting a new charge cycle is that this internal register can be written to. This means that the internal Growatt software can potentially modify this register any time it wants to during run time based on unknown criteria. For instance, at least one forum user speculated that the trigger "to resume charging is when battery voltage drops a set percent below float" (see
https://diysolarforum.com/threads/growatt-odd-behavior.22696/). While I do not see anything in the Growatt MODBUS pdf to support this theory, he may be right. So to be more sure of where this trigger is actually set during run time would require you to read the Growatt registers as discussed above (yes that is a pain).
This is all a guess at this point so do not take this information as truth! (everyone likes a disclaimer).