That's different- I don't know if it would be issue or not tying into a feed that eventually goes to battery. Other than controllers are not designed to supply load directly- being that its also going to battery, it might not.
But controller also 'reads' battery, possibly interfere with what controller thinks it needs to do? That and feed line, I assume, also goes thru the converters circuit board fuse for that line. Thru converter to get to the battery connections. Doesn't seem as direct as one might think. If fuse blew be like dumping by turning off disconnect.
Placing the controller behind in the refer box, unless it has remote temperature/battery sense for battery location, would see some temp variations that might affect output. That and access- many controllers have ability to shut off charge output. If controller disconnected from batteries long term, unless self powered, may need to reconfigure?
Too many unknown variables for me, may work just fine. Id make the effort to run directly to battery regardless how painful. Just my opinion.
If you do this, set up where panels output can be isolated from controller, switch/fused, and controller isolated from battery unless it had an off for output (so there is no chance for even minimal drain long term if disconnect left on). Then its developing a procedure/sequence of events to disconnect and reconnect solar and follow it. That sounds more painful to me than running wire...again just an opinion.