If I'm liveaboard travel diving, in expensive unknown waters in unknown conditions with unknown "buddies"and unknown guides I feel more comfortable having an SPG. I'm new to AI and have only 18 dives on mine. I had a DS4, which I know can be routed upside down, but I asked SW if having transmitter on left with computer on the right would cause issues, they replied saying they did not recommend this configuration. I did it anyway using a 6" hose, no signal issue at all. I've never in almost 20 years of diving had an SPG fail on me. I'm still learning to trust AI. You are a very experienced and knowledgeable diver, most are not, the OP has 15 dives. At my local dive site which I know better than the back of my hand I've been diving sans SPG, so far so good.
I have essentially no experience with AI though outside of a working knowledge of the technology. I'd personally run a DS4 upside-down in that scenario, though ironically if given the choice, I would run it upside-down anyway since it will have less of a chance to whack your head. I wouldn't put the transmitter on a short hose either unless you had "valet" service for your gear. I really don't like other people touching my gear, so I'd just try to avoid that to begin with, but that's just me.
Even at 15 logged dives though, you can set limits that mitigate the need for a backup SPG. Set the limit of, if it fails, and doesn't reconnect after say 3 minutes, then call the dive. Set two time limits, one based on a conservative sac rate and max depth=average depth, like the example above, and then have another one which is a pressure based limit.
In that scenario as an instructor going over dive planning with a student with AI, I'd do the following.
We require our students to calculate SAC rate for every dive during training and encourage them to do it on every dive thereafter. I tend to be a bit lazy about it because it is quite complicated when dealing with gas switching and decompression, but I do enough OW dives every year to keep track of it.
When setting the limits for that same 66ft dive, using NAUI tables with a 45min NDL because it's the first dive of the trip, we will go through the following.
Limits that you need are depth, time, and pressure.
Depth limit is 66ft, because that's your limit.
Time:
NDL=45 minutes, so you are limited by NDL to 45 minutes for this dive.
Gas:
Diving an AL80, at 66ft, and you recorded your SAC rate at 0.66cfm. Assuming that you stay at 66ft the whole time because it is a true square profile dive, you will be consuming 2cfm at depth. We know that your tank is 77.4cf, but we also know that regulators really need a 200psi buffer to function properly and 5 minute safety stop is what I like to plan for. 66+1.5+5=5cf=> 200psi+200psi for the regulators=>10cf. This gives us 67cf to use, divided by 2cfm yields your real time limit of 33 minutes. That gets you back on the boat with somewhere around 300-500psi give or take.
Gas trumps NDL in this case, so your time limit is 33 minutes.
Pressure: This is what normally turns most peoples first dives, and is a proper limit that can't be violated.
I believe in rock bottom calculations, and doing it with "Excited" sac rates, I'm not entirely a fan of the way that rock bottom is normally calculated. I do 3 minutes of conflict resolution at depth, 1cfm sac rates, and a 3 minute safety stop at 15 ft. 2cfm*3ata*3 minute=18cf for conflict resolution, 2cfm*1.5ata*3=9cf for safety stop, and I typically will leave the ascent blank because there is buffers built in there. 27cfm total=1050psi, plus 200psi for the regulator, and they have an ascent pressure of 1300psi. For al80's, my "rule" is 400psi/ata for down and dirty math. It's easy, and prevents having to calculate anything. That's my rule, and students have to calculate it.
Time round 2: If using AI especially, you have to estimate a time turnaround based on safety if you want to continue your dive in the event of failed pressure gauge.
We know that we are allowed to use 1700psi out of this AL80 based on the pressure limits above, and that 1700psi=44cf. From our previous SAC calculations, we can estimate that we will use 2cfm when at depth, and it should take us roughly 22 minutes to consume that amount of gas. We know that there is a real time limit of 33 minutes based on full gas consumption. To give us a bit of a buffer for safety, if we have an AI failure, as long as your bottom timer/computer/watch/buddy is functioning properly, then you have a contingency time limit of 20 minutes. No reason to stop the dive early so long as your watch works.
Aside from the AI bit, this is the type of planning that we teach at the basic open water level and is not beyond the abilities of someone with 15 dives. The difference is that at that level of their diving career, the limits are going to be exceptionally conservative because they haven't had enough time to really get a handle on their SAC rates and know what they are going to stabilize to. You are also adding a lot of conservatism by assuming your average depth=your max depth, but until you have the experience to know to change that, you have to use the conservative numbers.
If I am diving in a new cave for the first time, I use the max depth as my average depth for my time and decompression limits. One of my recent dives in Jackson Blue was 200 minutes with an average depth of 65 ft, but the max is 96. That's including deco, so for gas planning it's really an average of about 80ft. If I'm not familiar with that cave to know what the average depths are going to be, I have to assume it's flat at the bottom and stays that way forever. That means I'm probably planning a very long time with a lot of extra deco gas, as well as cutting myself short on the penetration limit that I set.
If I plan that dive, the planner says I have 51 minutes of planned deco. In reality, I had 39 minutes when I got to the bottom of the chimney. That means we planned well, had extra gas with us, and on the next time, we can either chose to adjust the gas plan to be a bit more accurate, or we can continue to plan a bit on the conservative side and have less to worry about on the way out.
I'm a lot more cavalier with my gas planning in OW, but the surface is accessible which makes all the difference in the world to me