When teaching rookies in wildcatting, their background makes a big difference.
For example, some folks hate running software and doing math, others don't mind it.
If you tend to be the type of person who can read software manuals, do their own taxes, and doesn't mind running personal computers... then you can learn to wildcat with internal ballistics software like Gordon's or QuickLoad.
Depending on experience, you should still study and compare as many published recipes as possible, even if they are not exactly the ones you plan to use. These other ones give you a preview of how the given powders and bullet weights played out and with experience you can anticipate which ones should help or which ones will be poor performers with the bullet you want to try.
Sometimes, when a cartridge is still a wildcat, new, or there isn't much published, you might find recipes that straddle what you are trying to find. For example, if there is data for a 68 grain, and a 73-75 grain, and you are trying for a 70 grain, you would be in decent shape by studying those published recipes. The warning is to be careful with non-traditional bullet construction which doesn't interpolate or extrapolate when compared to traditional bullets, for example monolithic bullets.
For example, take a look at the available online data and you will find the Hornady data is still up for free, and happens to show 68 grain and 73-75 grain data. It will be safe enough to study the start loads and work up from here if you are experienced in load development. Starting with your H4350 or StaBall 6.5, you can see there is little difference in the start levels between the weight ranges for these published recipes. To keep it simple, the H4350 would start at roughly 37.5 grains and the StaBall 6.5 would start at roughly 39.5. You should stop the H4350 at roughly 40.5, stop the StaBall 6.5 at roughly 44.5, or stop sooner if there are any pressure signs.
View attachment 8611159
So with the above, you could start up without too much risk, but if you wanted to reduce your risks further, you would run GRT/QL and see if the internal ballistics simulations gave any further insights. If any of this is too much risk for you, then stick to the paved paths.
YMMV