-> Summing up all of the above, I think the prospects are promising..
-> and worth trying.
I certainly wouldn't discourage TRYING it.  ANY or all facets of it...
But don't invest TOO much time effort or MONEY in any of this.
-> If anyone out there have experimented with any of these ideas, I
-> would like to hear about the results...
I've tinkered with most of these ideas at one time or the other over
the last thirty years, and I can tell you that while some of this is
marginally POSSIBLE, most of it is fiscally unfeasable.
Propane is a pretty good refrigerant, I use it myself in certain
applications, but BETTER than Freon?  No way.  Lots CHEAPER, especially
of late, but a system full of it is a potential bomb.  If you DO use it
be SURE it's VENTED!  FORCE ventalated if possible.
AS for EXTRA power from your PV array, if you've got enough to WASTE
that much you need more batteries to store it, thus increasing your
overall potential storage. OR you've got a lot more PVs than you NEED
for the loads you run.
As I said, these things CAN work, but practicality is somewhat lacking.
Believe me if there were some practical way of deriving cool from all
this extra sun *I* have, I'd be doing it. (Six miles from Mexico in DEEP
South Texas.)
 ^..^
--- FidoPCB v1.5 beta-'j'
---------------