Do you think you can use 10V at 100A or 1,000A?
If you are off the tables/curves then you use the last entry. This is common sense.
You can get a special spark-test done, if you wanted to use more current at your 3.9V, for example.
There are other limitations like Thermal considerations which will limit the current too, so it's not only ignition V/A curve then we can't get up to 100A or ..., i just see the curve which is going so much high when getting close to 10V !
However i didn't see anything about this matter in the standard, maybe you are right, the lower voltages should be tested by spark-test ... but i didn't find this note in the standard too, which i think should be mentioned ...
In fact i couldn't accept FVM comment, that's why i asked him the reason of his comment, (however no reply yet :-?) one of my reasons for not accepting it is:
There is another table-curve for limiting capacitance vs voltage, The table starts from 5V only.
according to FVM comment, the cap values for lower than 5V, will be same with the highest value below (for 5V).
There is an online free simple tool in exveritas website, the company is ATEX consultant and NB for certification, In the online tool you can enter the voltage and other factors to see the cap value (simply extracted from the standard table) but there is a note under the page :
"Upgrade to Intrinsically Pro for Inductance Assessment, Mixed Capacitance and Inductance Assessment, Capacitance levels below 5V and much more…."
The last sentence means the cap values for less than 5V is different than the highest value in table !