AC unit...
other things that may help:
better configuration for exhaust/intake
more efficient lights will produce less heat.
these things may be options for a few degrees but nothing major.
VPD is solid science... i don't dont the validity... i am curious how much research has been done to map out what is optimal and how a different species impacts those suggestions etc etc... until i read some white papers, hard to take the word of forums and websites that peddle all sorts of urban myths along with good science... or some bastardizations of science etc.
so, using hottest of your range -- 28c / 65 RH = 1.32 kPa. This is fine for flower, maybe a tad high for vege, supposedly..
lets say running dehum brings you to 30c / 55% RH, which is a safer RH range to avoid microbials. That is a 1.91 Kpa... probably less than optimal for vege phase, and a little out of range for flower. I had a VPD of 2.5 and it didn't kill anything... i did have to reduce feed significantly, though. the second run was still high, but much better and the results improved. So, i'd guess you'd have some slowdown of some sort with vege'ing plants... though does that translate for autoflowers too? cause they don't change light schedule, so i hesitate to think this is a parallel for them.
You can mitigate a high VPD with a slightly lower concentration in your nutrient mix... sure the plant is still working a bit harder than normal taking in water, but you can ensure it still takes in same grams per day of N/P/K/et al. This won't solve all problems, but will avoid the plant building up nutes beyond what it can metabolize in 1 day due to increased daily drinking caused by elevated VPD.
obviously, a plant is growing so it needs a bit more each day, but if you have been giving a gallon a day, then after environment immediately changes it needs 1.25 gallons per day... then you know it's probably a good idea to reduce concentration to match new rate of drinking cause no overnight growth will cause a 25% increase.
this is just one more relativistic factor that impacts everything else, or just about everything else. this is why we can feed "exactly the same" and get different results. (just one reason and not limited to this 1)