Since "Salary" is just one component of what players can make, but you would expect "Payroll" to really be the amount paid to all players in a given year, the two numbers should be different, although the Pats could be seventh in both categories independently.
As to the question of whether we spent 94 mil on a 92 mil cap, that might be the case, if you calculate it including incentives paid, which (if I understand this right,) reduces your team's adjusted cap for the following season. I'm pretty murky on this particular point, and you need a real capologist to get it right.
Last point... where you rank in payroll is significant, only in that you should spend to the cap every year, in my opinion, one way or another. If you miss that goal it shouldn't be an ongoing thing. And you should do it in such a way that you get max value from your spending.
Where you rank in salary is insignificant. Assuming you're spending to the cap, your salary spending only indicates a philosophy of paying in salary and keeping bonuses down as a total proportion. If you can use bonuses to the mutual advantage of the players and the team, you could very easily end up with a low salary total, because a lot of salary is going into bonuses.
The only other tea leaf to read into these proportions is the question of incentive tied to big bonuses. That is, you don't pay a bonus you'll be prorating for five years, if a player is injury prone or not otherwise motivated... those bonuses take money off the table and into the player's pocket, with no guarantee of performance (except to the extent that the bonuses are tied to incentives.) On the other hand, bonuses can move cap hits from year A to years B,C,D, etc... or could put one big hit in year A, and keep salaries low in B,C,D... whichever direction you desire the accounting to go. So the available tools can easily affect whether you have a "high" or "low" rank in salary in a given year, without changing how much each player gets.
PFnV