What you propose will work in practice in many cases but has some potential problems for some situations.
NimH charging relies in approximate order of use one or more of the following for end of charge detection:
negative delta V
delta temperature
absolute temperature
terminal voltage
timer
All of these may be defeated to a greater or lesser extent in systems that have power turned on or off during the charge cycle. Degree of effect depends on equipment, position in charge cycle, ambient temperature and more. YMMV and probably will.
NimH "like" to be charged at about 1C rate to get a clear negative delta signal. Lower rates tend to mask this signal as it gets smaller or vanishes. Turning charging on an off around EOC is liable to make this unreliable.
Delta temperature MAY be useless if the device is charged intermittently.
Absolute temperature likewise.
Terminal voltage "settles down" given rest periods. It will come back to "correct state" fairly rapidly but few chargers probably use it.
If a timer rests repeatedly it too fails.
So. it seems likely taht a NimH charger may have "a hard time" depending in which combination of EOC techniques are used.
Lithium Ion will be well behaved in most cases.
NiCd has similar problems to NimH
Lead acid should be OK for main charge aspects but topping cycles and the like would get greatly messed up.
Targets have to be able to withstand peak voltage - easily established bt fatal in some cases if you get it wrong.
If load Watts of all attached devices approach panel Watts in high sun, multiple clients may oscillate / interact over seconds to munutes in low sun as appliances come on line, start charging, load drops voltage and som devices drop out rater than throttle back.
Many LiIon chargers want the energy source to be able to meet their current demands. If the source sags under the demand they will stop completely rather than backing off. eg a laptop that wants 3.5A will usually not charge at all if only 2A is available.