In comparing the two graphs above, it's easy to see that in the top graph (tournaments), the only wild swings happen when I manage to hit big in a tourney, then there is the gradual downward trend until the next hit. And then there's the bottom graph...
The cash game graph starts out with a mild downswing that gets more pronounced, then there is a large upswing from when I was playing well and getting good situations, with an almost vertical rise before it nearly crashes back to break-even in shark tooth fashion. What that means is I took shots and some higher buyin levels like $1\2 (risking $200). Naturally it's a rush when the game is running your way, but a total bummer when the downswings hit.
Sometimes the downswings were due to solid play on the flop, where I got all the chips in ahead and simply got outdrawn on the turn or the river. It's part of the game that you have to embrace. But other times it's from getting outplayed on the turn. That's a leak that I'll be working on, but I'll be doing it at .10\.25 (risking $25) and below.
Even though my overall poker bankroll is well above $1000, I want to cap it online at what I've managed to earn there, and grind it up gradually. I'd be far better off now if I had dropped back down after the large upswing, but that's a lesson all poker players learn. You never really quit while you're ahead, you just play at reasonable levels so that you don't get rattled when the next downswing happens, whether from variance or from your game unraveling.
Tournaments have been my bread and butter in this experiment, so I will keep playing those at the $10 buyin and below, and keep learning cash game play at the micro-levels. It really sucked to drop below $1000 a few months ago, and now that I'm above that threshold again, I'd like to keep it moving up gradually.
It takes a lot of discipline not to hop on the $1\2 tables to try and win a monster pot, but I think I've learned my lesson about bankroll management in the cash games.