The first problem is the draft glitch, where the CPU only takes QBs, Ts, DEs, and HBs in the 1st round.
Solution: Turn on user control for every team prior to the draft. Manually select 1st round picks for every team based on BPA and team needs. Save the file, and make the other teams CPU controlled again. Reload the franchise save and complete the draft. The rest of the draft will continue without a hitch, the CPU will have appropriately selected first rounders, and you will be very familiar with every single team's first pick. W/o making this adjustment, selecting QBs, DEs, Ts, and HBs after the first round will give you very poorly rated players as all the good to decent ones were snatched up early.
The second problem is not one I'm familiar with. After hundreds of hours playing this game, I actually think the players coming out of the 2k draft are pretty good. For starters, the player ratings right after the draft usually aren't what they are when the season actually starts. I drafted a guard one year in the 3rd round. After the draft he was a 78. After preseason, on the opening day, he was all the way up to an 88. So wait until after the preseason to see how the rookies really are.
Secondly, make sure you're using the rookie report to see how players are ranked. You can exit the draft at any time to see the rookie report in the ESPN.com section. It is kind of a pain to exit the draft just to check how players are rated, but it makes things a lot easier since inside the draft all the rookies are just listed alphabetically, unlike in 2k4 where they were listed by ranking. If you try to balance taking the BPA and addressing team needs, you should end up with a very solid draft. Most rookies except for are rated between 60-80 on the game...unless you choose players that are really far down on the rankings.
Another thing you can do is look at who has already been drafted. You can view the ratings of rookies just selected by using the right thumbstick click, while looking at the selection history. If you're torn between taking an OLB and a C, and you see that the last OLB taken was rated an 80 and the last C taken was a 65, you can make an educated guess that the OLB and C just under those selections in the rookie report rankings will be just below them. I've used this method to get a decent estimate of what a player's potential rating might be.
Hope this helps. Trust me, if you pay attention to the rookie report while drafting, pay attention to the picks the CPU makes, and don't reach for positions of need, you can get some really good finds late in the draft. If you put time into the draft, you can find yourself finding a 70 rated QB in the 6th round rather than a 25 rated QB in the 3rd. Good luck.