Ok, I can verify running Mitsubishi mini splits for five years no problems; also if something were to go bad it would probably be the air handler which can be cheaply replaced; the outside units are robust. Get a Mitsubishi, Fujitsu or Friedrich.
On the Chillers vs. ACs I must disagree that chillers are more efficient much less better. Not to be disagreeable but in the search for knowledge lets dispel some myths:
First, chillers are in fact modified ACs, they use the exact same technology and work in the same way but instead of cooling the air with an air handler they cool water and or coolant with a pump. So if an AC's performance is negatively effected by humidity or altitude than so will the chiller's.
Second, chillers are like old low tech ACs in that they don't use inverters like the latest most efficient ACs. Meaning they don't modulate their output according to what is needed they either run or they don't and as many know they consume the most power starting up.
Third, when looking at power consumption you have to take into account that chillers work in conjunction with fans and pumps which consume power. So when calculating for purposes of comparing cooling setups lets agree to add up all the power consumed in cooling a given grow room.
Fourth, with chillers you are using a condenser and fan to extract heat from liquid via refrigerant and then in turn using the cooled liquid to cool your grow room air. Essentially adding a step in the heat transference process which is not efficient; with an AC you extract heat from you room via refrigerant and then expel that heat via outside air with a fan and condenser.
I believe that a new inverter style AC will be more efficient than a chiller system.
PS: Although I obviously don't think I am, I could be wrong about this and am curious to see what others will say. I keep an open mind.
I'll address your points in order:
1. You've got the wrong end when considering the efficiency deficiency of AC vs. chiller; it's the ambient air that impacts an AC unit's efficiency, that is, the fact that the air the AC unit is trying to cool isn't very thermally dense. The fan blowing through the compressor section isn't the problem. It's trying to get the COLD from the AC unit to its destination that gets tougher as altitude climbs and humidity falls.
2. True. But since when is running at half speed always more efficient? A chiller can get away with running at 100% or not at all because the water it's cooling is the cold reservoir for the system. When the water warms up past the chiller's setpoint, it runs at 100% output, AND at its best efficiency setting, until the water is cooled. It then shuts off completely until the water is again too warm. The length of time it spends idle does indeed contribute directly- and mightily- to its efficient performance. Finally, the excess power consumed at startup lasts for less than a whole second, not very long when considering a given operating cycle lasts minutes.
3. Okay, that's fair. Let's also remember to add in all fans used to blow cold air around from a minisplit.
4. You're right in that the heat energy is moved from the growroom air and put into water. What you're missing is the efficiency at which this process takes place. Because of water's much greater thermal density, the efficiency at which heat is moved from one media to another is much more important to the overall calculation than merely whether or how often it happens. The water is a thermal reservoir. It is here where 'cold' is stored for use in the growroom while the chiller is off.
Since we don't have the same growing setups, it's tough to compare apples to apples. ChillKing's 2 Ton chiller unit runs on 240v and pulls 8.5 amps. That's roughly 2000 watts- when it's running. When it's not running, it's effectively zero. Yes, there are pumps and such still running, but they don't add up to more than 100 watts. Even when cooling 8kW of sealed and vented hoods, 3 RDWC systems and an Ice Box in another room all simultaneously, it doesn't run constantly. It runs about 2/3 of the time... Best of all (and this is why I chose this style in the first place) I will not need to buy another AC unit to cool a second room on the flip- I just plumb a few more Ice Boxes into the existing system and plug the fans that blow through them into another temp. controller and I'm done! The chiller will keep up with the second room with no trouble, as its heat load occurs only when the first room is off.
As far as complexity goes, I would much rather maintain one chiller than 3 different AC units! By the way, what are the wattage outputs of all of those systems?
Let us also not forget that I have yet to see an AC unit that chills the water in a hydro system.
Finally, to use outside examples; architects and engineers design modern buildings to be as energy efficient as possible, knowing that the operating cost of cooling a building will quickly outpace whatever installation cost differences might be associated with one system vs. another. Most of these buildings use water chilling, for all the reasons I've outlined above. If AC were truly more efficient, you can bet they would use it instead!
Nowhere have I said that a chiller is 'in all cases better no matter what' than other approaches. I'm told heat pumps are very effective in high humidity situations, often operating above 100% efficiency. I don't know anything more than what I've been told in this regard. Also, AC can be very effective in low altitude/high humidity applications because it can reduce RH. So can a chiller, but the performance difference isn't as great.