Don't give up so easily!
The easy way to simulate the effect is to take a composition and run it through CalcZAF with different takeoff angles. It's not exactly the same geometry as your tilted sample, but similar enough.
But before we do that, remember, you can rotate the tilted sample in the sample holder to align it so that for two opposing spectrometers, the take-off angle is close to 40 degrees. It would be sensible to run your major elements on these two spectrometers and use the other spectrometers, which will essentially have smaller or larger take off angles but won't show the tilt effects as obviously, for trace elements.
Since x-rays with similar energies (and mass absorption coefficients) won't be affected very differently by different take off angles, let's pick a composition with a range of emission line energies, say, a sodium silicate glass. Here's the first calculation from CalcZAF at 40 degrees:
ELEMENT ABSCOR FLUCOR ZEDCOR ZAFCOR STP-POW BKS-COR F(x)u Ec Eo/Ec MACs
Fe ka 1.0123 1.0000 1.1033 1.1169 1.1450 .9636 .9727 7.1120 2.1091 126.947
Ca ka 1.0432 .9897 .9903 1.0224 1.0078 .9826 .9230 4.0390 3.7138 310.118
Si ka 1.4496 .9977 .9581 1.3857 .9345 1.0253 .6247 1.8390 8.1566 1785.49
Al ka 1.5253 .9925 .9840 1.4896 .9522 1.0333 .5818 1.5600 9.6154 2059.83
Na ka 2.2531 .9967 .9724 2.1838 .9280 1.0479 .3736 1.0730 13.9795 4105.93
O ka 3.8333 .9984 .8922 3.4146 .8396 1.0626 .1828 .5317 28.2114 8514.37
We'll want to watch ABSCOR values as that is the absorption correction and therefore the major player for tilted samples since we are essentially changing the absorption path length by tilting to other angles. Basically, the lower the takeoff angle, the greater the absorption path length and the greater the absorption effect.
Now, by using the Analytical | Operating Conditions menu we can try other take off angles. Here is 30 degrees:
ELEMENT ABSCOR FLUCOR ZEDCOR ZAFCOR STP-POW BKS-COR F(x)u Ec Eo/Ec MACs
Fe ka 1.0157 1.0000 1.1033 1.1207 1.1450 .9636 .9651 7.1120 2.1091 126.947
Ca ka 1.0554 .9907 .9903 1.0354 1.0078 .9826 .9027 4.0390 3.7138 310.118
Si ka 1.5832 .9979 .9581 1.5137 .9345 1.0253 .5565 1.8390 8.1566 1785.49
Al ka 1.6811 .9929 .9840 1.6424 .9522 1.0333 .5107 1.5600 9.6154 2059.83
Na ka 2.6354 .9969 .9724 2.5548 .9280 1.0479 .3048 1.0730 13.9795 4105.93
O ka 4.5934 .9985 .8922 4.0920 .8396 1.0626 .1392 .5317 28.2114 8514.37
And here is 50 degrees:
ELEMENT ABSCOR FLUCOR ZEDCOR ZAFCOR STP-POW BKS-COR F(x)u Ec Eo/Ec MACs
Fe ka 1.0103 1.0000 1.1033 1.1147 1.1450 .9636 .9770 7.1120 2.1091 126.947
Ca ka 1.0363 .9890 .9903 1.0149 1.0078 .9826 .9348 4.0390 3.7138 310.118
Si ka 1.3750 .9976 .9581 1.3142 .9345 1.0253 .6690 1.8390 8.1566 1785.49
Al ka 1.4381 .9922 .9840 1.4041 .9522 1.0333 .6287 1.5600 9.6154 2059.83
Na ka 2.0388 .9966 .9724 1.9758 .9280 1.0479 .4241 1.0730 13.9795 4105.93
O ka 3.3771 .9984 .8922 3.0080 .8396 1.0626 .2189 .5317 28.2114 8514.37
So clearly for lower energy elements, this is a significant concern, but for higher energy x-rays such as Fe ka, the effect is much smaller and might even be ignored.