I cater the Spire to the top each week. I'm in Chapter 18 but started when the Spire started...probably Chapter 13-15 somewhere. The cost is high but not so high if you work the logic of the chests carefully. My methods result in my spending about 1.2 times more than the actual costs if I hit every chest in three tries. That and about 125-150 diamonds which make back from my fs silver level and the occasional winnings in the Spire itself.
In general when I cater I work from the left to the right in the first round, matching one good to each guard and rotating to the front (far left, coins) if there are more guards than goods. I then count the satisfied guards (green checked) and give them one point. Then I look for the "not needed" -- the greyed out goods in the goods I have to offer -- and they get two points. If the number of points from these two sources is more than or equal to the number of goods I originally had to offer, I continue. Less, well, sometimes I continue, but most of the time not. I start again. If you have just 3 goods to offer at the beginning you should never start over unless you accidentally offered the same good to the same guard twice. But even you can usually win in the third round. You can almost always win the 4 goods level if you just load your choices from the left AND, check before you try to see that you haven't offered the same thing to the same guard as you did in the first round. Usually with 4 goods the third round comes up with two or three "wrong person" and that's almost always something you can figure out. On occassion you'll get two "wrong person" and one "not needed." Process of elimination will fill two of the three and often leave you with the third guard having two possiblities. Count the number of times each of the two has been accepted. The lowest is the one you use. If they are tied? Offer the one the farthest to the left...usually coins or supplies.
All the above applies to starting with 5 goods to offer so long as you remember that after the first round you will need the points from the "not needed" and "satisfied guards" to add up to 5 -- the number of goods you have to offer at the beginning.
Ditto for 6 goods, though the "success" rate drops a bit and you will need to carefully consider if you should continue if the count is 5. Ideally it should be 6 or above. 7 should be 7 or above, but since that takes a lot of roles you may need to use the free thing for one guard or spend 25 coins for an extra try. I usually save the 25 for the two top bosses and when I try more than twice on one of the third tier 6 goods chests.
In the end, I believe, the first two levels of the Spire are pretty easily won so long as you use logic and are systematic. The third level is harder, but not insurmountable so long as you are willing to spend another 20% to start over once about 1/2 of the time when doing 6 goods chests (i.e. everything but the gates and the boss), and spend 25-75 diamonds.
That's how I cater the spire to the top. There are details about choices to make and the logic which I did not include but it works for me.
AJ