Has anyone figured out the centenary curve on a long but taut string? (I haven't.) It could be minimized by using some of that high tech fishing line that's very strong for its size and using a fair amount of tension. If the catenary curve was still significant, it could still be calculated and allowed for. I think, using a very thick foam core or deep truss, you could make a table that stayed flat when moved. I'll admit I haven't calculated how deep such a table would need to be. If you had access to a flat table for a couple of days, you could build yours on top of it, saving a bunch of trouble. I wonder how flat that pond is when it freezes? If it's flat enough, find some glue that sets when cold, and make your table on that.