Hmm, so the DIY was too weak as compared to OEM with the batch you made. The larger volumes you are using will more easily show discrepancies in the ratio, than the smaller batch the previous poster was using. Measurement errors can swing the results more easily, both positively and negatively, when the volumes being measured and mixed are small.I dicided to give the DIY reagent a go tonight by measuring by weight rather than using measuring cylinders.
I also decided to mix it on a large scale but still following the ratio given above.
5 litres of 0.1N mixed with 16.2 litres of Rodi water. Totalling 21.2 litres of reagent. Unfortunately I think the ratio given above isn't completely accurate for larger scale batches.
Xepta reagent test: 8.00dkh
Diy reagent test: 8.99dkh
A 1:3 ratio would be closer to matching based on your results.