The roles don't matter, it is the usage rules that have demolished any semblance of a bullpen usage availability equal to the amount of innings the game implies you should allow. I have had Dizzy Dean with no roles come into 7 games in a year resulting in his pitching 328 innings and apparently that is all Kosher. Put in a bunch of 9L and 9R guys with him have good starters with an R in their rating, which are much cheaper than expensive bullpen pitchers and besides that can start and you can save a lot of money on the bullpen in lower league levels.
The game continues to devolve less as a baseball simulation and more and more into a contrived computerized valuation modeling exercise. As valuation models take over, competing baseball strategies become uncompetitive because the vast value modeling capabilities overwhelm any strategy implementation. Another role for relief pitchers is not going to change the destruction of the game that is evolving as a result of the massive valuation change. That Kent Tekulve is one point one million dollar salary and is far more productive against right handers than the 7.7 million dollar relief villians Bruce Sutter and Dale Murray is a mystery to me as to what the pricing model is trying to accomplish, but Tekulve's ilk is the type a park and hitter valuation model can impact.
The best teams even in 140 million dollar leagues are built on the value of 4 9 inning starters under the most expensive range, because their value is better than other ranges and you can get 103 wins with less than 40 innings of bullpen in our simulation now.
https://365.strat-o-matic.com/team/sim/1595811Adding another role to the bullpen is not going to have any impact on the game and is equivalent to adding another attractive magician's assistant to a magic show.
Personally I would like to see a limit of 12 imposed on the total bias of the composition of teams for both pitchers and hitters to eliminate this valuation strategy. This would make the drafting of Mariano Duncan and Tekulve as needed as part of a strategy of the game and less of a valuation tool based on the ballpark effects.