I'm planning a small bench (picture a piano bench) and am doing "cabriole" legs as an excuse to do more spindle turning and to try something new. My latest test piece is pretty decent but I'm a bit vexxed by getting the convergence point to end right below the pommel.

I've been referencing these two articles:

Jon Siegel
Ernie Conover

If you manage to parse the exact processes described in the articles, you'll noticed that they each start at a different axis. Jon on the offset, Ernie on the normal axis. Assuming I'm reading it right, Jon says to cut the pommel's bottom on the offset axis whereas Ernie says to make that skew cut on the main axis.

In my test piece I followed Jon's process, making that cut first on the offset. I then later cleaned it up on the main axis to make it symmetrical. I was then left with the tapered section kind of awkwardly meeting the circular section so I added a cove to try and hide it.. bad idea, it made it stand out even more!

So I guess I'm looking for more input on processes that work well for this project. And also, is there a way to determine where the convergence point is when the billet is on the lathe? There are a lot of errors introduced in measuring and punching holes.. there's no way the actual convergence point is going to be right where I marked.

Test piece: