Sorry if I'm wearing you out with the long posts, but, well, EFX can be complicated.
But back to reverb and chorus:
To get a patch to sound in performance *exactly* the way it sounds in patch mode, you also have to set up the reverb and chorus *types* to be the same and then the routing and send levels to be the same.
Probably the easiest way to do this is simply to use Roland's copy method (see previous post).
The problem I find is that although now the patch does indeed sound the same as in patch mode . . . what about the other 15 players in the ensemble? What are the chances that those 15 players are going to require exactly those types of reverb and chorus with exactly those general levels to achieve the best (and here's the key phrase) *overall mix* for your performance and for the song?
They might, but I find generally that imposing a patch's patch-mode chorus and reverb settings on *all* the other instruments rarely works. Yes, you can tweak the send levels for each perf. part, but you've still had that one patch choose the reverb and chorus types and general settings.
My philosophy is: "You have but one chorus type and one reverb type for *everyone* in your band (=performance). Choose wisely."
So I sacrifice the reverb/chorus types and general settings of that one patch whose [insert] EFX are supposed to sound exactly as they do in patch mode.
Instead, I choose the reverb/chorus types and general settings that best suit the overall performance and tune, then tweak that one patch the best I can to duplicate its rev/chor sound in patch mode. You can get closer than you might think. After all, in patch mode, that patch sounds one way alone; in perf. mode, it's in an ensemble setting and probably needs different rev/chor settings anyway.
This is also why I don't use the "copy" method of copying a patch's EFX settings into a performance part.
So: I set up my EFX Src for the performance to a certain performance part (as explained above; my part is 15) and then set my performance rev/chor types and general settings independently for the performance itself, and finally then tweak the rev/chor send levels for each part in my performance according to what each part/instrument needs.
You can tweak the part send levels at the beginning of the song using CC91 and CC93, or set them manually and save them as part of the performance, then use CC91 and CC93 during the tune if you want to change them.
There are a couple of caveats with setting up your performance to make sure you have enough headroom for your rev/chor levels. If anyone is interested, we can look at that, too.
But I'm sure you're about "read out" by now.
So play with this stuff a bit and see if it helps.
Good luck.