2. As a general approach, the successful teams generally seem to be the ones that develop their core players through the draft. The Patriots, Ravens, Steelers, Packers and Seahawks all fall into that category. For the Patriots, almost all of the key players - the ones that you want to build around long term and really don't want to let get away - were developed through the draft: Tom Brady, Rob Gronkowski, Jerod Mayo, Devin McCourty, Chandler Jones, Dont'a Hightower, Jamie Collins, Nate Solder and Sebastian Vollmer currently, and guys like Vince Wilfork, Richard Seymour, Logan Mankins and (shudder) Aaron Hernandez in the past. This has the advantage that you don't throw big money at players until they are known quantities with proven ability to perform in your system. In the case of Hernandez, the calculus went disastrously wrong because of off-field issues.
3. Attempting to fill acute needs through the draft is generally in opposition with building a strong talent base through the draft. It leads to reaching to fill specific positions, and passing up more talented options. The more freedom you have to move around in the draft and to not have to fill specific needs, the more you can focus on building a talent base for the long term. The draft also has a sizable inherent failure rate, so diversification and stockpiling of talent makes sense as a general approach. Some successful teams explicitly attempt to factor this uncertainty in their draft approach: BB by trading down, the Ravens by accumulating extra comp picks.
4. In order to not use the draft to address acute needs, free agency comes into play. Upgrading talent through free agency is much riskier than plugging holes, because you generally pay a premium to upgrade talent. There are 2 problems with this: (1) you are to a large extent paying for past performance with another team, under different circumstances; and (2) it is difficult to extrapolate this to future performance in a new situation, where failure can be very costly. So unless you are very certain that you are getting something special or are getting it at a significant discount (e.g., Revis in 2014, the Moss trade in 2007, even though that wasn't strictly a FA situation) it's risky to spend on big name FAs.
5. As a theory, I would argue that the vast majority of players in the NFL are eminently replaceable, and are separated in ability by only a small amount. Only a small percent are the rare difference makers who would likely succeed wherever they play. The majority of those that have been productive in a past situation often were in a good environment to succeed, and may struggle to duplicate that success elsewhere. Many of them get ridiculously overpaid in free agency. The cardinal sin in free agency is to pay excessively for a player that's only marginally better than someone you could get for a fraction of the cost.