That inspection time was approximately 30 seconds long.
That's a VERY GOOD THING.
You need to practice using your inspection time + cross, then inspection time + cross + F2L, when you're good enough. But for now, you should really focus on fully planning out the cross and then executing it. Don't worry about how long it takes to do plan it until you're ready to speed that up.
But that's not my main advice. Rather, I wanted to make a comment on the mental aspects of practice.
Depending on the situation in which brain functions best, you can take all of the advice from the forum and try to improve in several areas at once (if you're just that awesome), or, if you progress very slowly (like me), then you'll want to only take on one thing at a time, master that one thing, then go on to the next. It's hard for me to focus on 15 different important things at the same time AND do well.
Then again, everybody probably does it the latter way (i.e. one thing at a time); they just do it WAAAYYY faster than me. I'm very slow. (As in brain. Very slow-brained.) I also don't organize well, so I tend to take too many things on at once without intending to do so. That's probably why it's taken me 5 years as opposed to 3 or less years to get to where I am now.
(5 years, and I'm only averaging 12-13 seconds? weeeak.)
In short, just make sure you know how your brain functions. Organization and planned practice, I think, is key to speedy improvement.
(i.e. Open on Microsoft Word Document, or maybe an iCal document, make a schedule, follow it. "This week, I'll try to master/improve blah-blah, and next week, I'll try to master/improve blah-blah-blah. I'll spend a day or two on blah because I don't think it'll take too long to improve in that area. I should spend some time on finger tricks every day. This weekend, I'll learn about 5-10 OLL/PLL algorithms and get used to applying them. Next month, I'll work on big cubes more often. The next month, I'll practice 3x3 again. For the first half of the month, I'll practice incorporating edge control into my solves. In the 2nd half of the month, I'll try to learn a new method, like Roux, or Petrus, or that dreadfully-confusing Ryan Heise Method. Maybe I should try learning COLL too, or F2LL + Winter Variation next month. I should figure out a plan on how I can keep using them so I don't forget all those algorithms... OH SHOOT! I forgot about 2x2!! I should probably start with ortega and guimond before I learn CLL + Anti-CLL. I don't wanna kill myself with too many algorithms. Gosh, so much to do. I better stay organized, or I'll never learn all this crap... You know, I didn't reserve a time for practicing 3x3x3 blindfold. I don't even know which method to use! Looks like it's time to do some research!")
Some weird stream-of-consciousness of the sort above this sentence should be going on in your head when making a schedule.
