One of my lessons learned is to do the testing on the lure before you finsh it.  Go somewhare that you can get at least 1/2 full cast length to be able to evaluate how the lure will run.  Tub tests and short tosses don't give you enough feedback.  If you are unhappy with the results you can tweak the bait before you spend a lot of time on the finish work..

 After forming the lure and installing the hardware, I seal coat the wood with 2T epoxy before doing any paint. This allows me to test the lure in any open water that I can find this time of year. There is turbine discharge below a local power dam that stays open where I can toss lures and evaluate. It is not too big a deal to add or drill out some weight at this point in time for fine tuning. Keep in mind that you will always loose bouyancy after paint and final coat. I have yet to make a lure that came out "too light" and almost always come out heavier than I was shooting for.

 Also, try to limit your work to the same pattern. I spent a lot of time making different shapes and going through the inventing process over and over again. I was less than 50% in terms of lures that I liked the final action, and lures that left something to be desired. Last year I stuck to a few proven designs and my results were much better although I still have left room to improve...

 Finally - keep good notes on your recipe for each lure design so that you can make it the same way next time. Placement of weight, amount of weight, depth of holes that you drill, etc. I use poster board templates with drawings of where everythign goes and how many lead sinkers I melt into each location. I like the lexan idea mentioned above even better than poster board (more durable).

IMO, crank baits are probably the toughtest to make because of their aggressive diving characteristic and potential for thigns to go awry.  Twitch baits are easiest because you need only worry about adding enought weight to stabilize and can cheat by keeping the weight forward of the "center of bouyancy".  Glide baits are somewhere in the middle because the "center of bouyancy" is very important, and it is very difficult to hit the near neutral bouyancy that you want for this type of lure.

When I describe the "center of bouyancy" this is the spot where you can grab the lure with a finger on each side and sink it with it's back level to the horizon.  I usually mark this spot and add lead here to keep it level.