We've recently developed a projection app for comparing players while factoring in Over/Under, Spread, Opponent, Home/Away, Salary, and Expected Minutes. Much of this post will rely on the projections made by this app.
Toronto @ Philadelphia (Unkown: Giving my best guess at O/U 210, TOR -2)
Milwaukee @ Washington (O/U 214.5, WAS -5)
New York @ Brooklyn (O/U 209.5, BKN -2.5)
San Antonio @ Atlanta (O/U 205, SA -7)
Miami @ Chicago (O/U 207, CHI -1)
Golden State @ Cleveland (O/U 230, GS -5.5)
Sacramento @ Oklahoma City (O/U 209.5, OKC -11.5)
Indiana @ Utah (?)
Houston @ LA Clippers (?)
The projection likes Simmons as a 2-to-1 favorite to be the best value of these three point guards, even at his $8.4K salary. Jack seems unlikely to be the best value if the projection is not scaled to minutes.
However, Jack and Dinwiddie have both been receiving increased minutes as of late. If we expect them to play 34 minutes (which has been their average over the last 10 games), and scale Simmons to his season average of 35 minutes, the projections are a little more even, with Dinwiddie actually having a slight edge over Simmons.
All things considered, I'm not terribly high on Jack, who has been playing very well lately, because I expect his usage to trend down as Tim Hardaway's involvement increases. Simmons seems like a good play if you can afford him, but Dinwiddie is a nice option if the NYK-BKN game remains competitive (which it should).
The projection model is not high on DeRozan, but it should be noted that they model doesn't account for "the hot hand", which DeRozan has had early in 2018. In a shooting guard field that is rather thin, I don't think DeRozan is unplayable, but as the model suggests, there's likely to be better value out there.
If you are going to play DeRozan, he should almost certainly be played at SG, because there is way more premiere value at small forward. According to the model, Giannis is the most likely of himself, Simmons, and DeRozan to have the best per-$ value. However, his upside is capped due to his high salary, while Simmons at a lower salary has a little more per-$ upside.
In further support of Giannis, his value holds up against some high-end power forwards. It should be noted that the model can't yet model for Kawhi Leonard because there hasn't been enough data accumulation for him, but I also think he has good upside because of the big defensive numbers he can produce. Porzingis doesn't seem to be a bad option if you're not willing to splurge on Giannis, as his value upside is comparable to that of Giannis, but his most-likely projection is a lower value than that of Giannis.
Porzingis is a little more playable at center, where he has more than a 40% chance of being the top center among the above trio of high-end centers. Aldridge could also be in the mix, but I'm concerned that his usage could trend down with Kawhi Leonard fully expected to play. As much upside as Embiid has, his price is just too high to warrant a ton of confidence in him at center.