Interesting stuff about the roof. I didn't know all of that about the paper being the key.
I had an asphalt shingle roof, and the screws went through the shingles, through the paper, and through the underlying wood and entered into the space of the attic. I never had leaks, but there were many leak horror stories out there.
I would be less comfortable with mounting hardware penetrating paper and sheathing, for sure. Though that's all designed for that. Think about it, asphalt shingles are stapled or nailed on, right thought the shingle, the paper and the wood. The shingles and paper both "heal" around the shaft of the fastener, and make it water tight. Same would happen with other types of fasteners.
Also with high winds there are stories of folks sustaining roof damage from wind creating lift under the panels and ripping the screws and mounting out of the roof. With heliocol panels this isn't as much of a concern because of the space between the tubes, but not all panels are made that way.
Absolutely. I'm still sweating that out a bit, because I didn't penetrate the sheathing or framing, and we get some good winds. But I installed mine as the solar company installs them, and they coached me. So I'm rolling the dice my panels will meet the stress. The Heliocol's have tubes, but also a unique bracket system that together address their wind bearing capabilities.
Our panels (Solar Industries) were more like a mat, and it seemed that they might be to create lift under the right conditions, which could seemingly be offset with a proper installation. The upside of this type of panel was better heating performance on Category B and C days, when compared to a heliocol panel.
I didn't like that mat design for the wind reason. I didn't research about the performance as much as I should have. I know that Heliocols are up there, but I wouldn't be surprised to learn there are better brands. And keep in mind, a lot of what I (we?) know about all this is based in no small part on the hype and marketing of the installers and the manufacturers. So who knows?! I think there is independent data for this, but that could be subject to interpretation, too.
So between the screws going all the way through the shingles, paper, and wood, and the stress load of wind on the panels, over the years, seemed reasonable to me to think this might be a little bit more work for a roof than simply sitting there with no screws and no additional wind stress, maybe not though.
That's logical, I wouldn't argue that. But my panels are attached to the tiles, and unless there's enough lift, the tiles just absorb the movement. Mine are not actually attached to the roof, they're just sort of stacked on each other, so my roof doesn't get all that stress. Additionally, I learned from my PV installer (heavier, more wind-prone panels), that a typical roof is over built to accommodate such things, and the additional stress and weight of solar panels still spec well under the roof's capabilities. This is assuming you have a roof that was built to such a spec. I know mine was, YMMV. That is also assuming that that wasn't more installer hype!!
Like I said, I think pool solar was the best home improvement we ever did, but almost everything has pros and cons.
As for the pump/vacuum issue, you want your vacuum and pump to run at night for safety, but with solar don't you now have to run the pump during the most of the daylight hours too, for you that would seem to be a "con", no?
Yes and no. Running solar "costs" flow, sure. And solar impacts flow in a different way than a NG heater, but only because heaters are normally plumbed to be inline all the time. I was taught here that my NG heater impacts flow by up to 30%. That's a lot! If I plumbed my setup to bypass my NG heater like I do my solar, then it'd have the same impact on the vac issue that my solar does. See? So technically that's a con for an NG heater, which is sucking flow 24/7, whether it's heating or not. My solar only does so when it's heating, because it gets bypassed when not in use. The vac issue is not germane. It mucks things up equally, solar or NG, if NG was being bypassed as solar is. This is about solar cons compared to NG (the OPs original question).
And you're running the pump at flow rates enough to power your vacuum, and in the day enough to optimize the solar, so what kind of flow rates do you use for that? I'm not trying to be nasty, and I can understand if it sounds like I am-- I'm a bit awkward socially, but I really like to know things like flow rates-- I wish I had gotten an intelliflo pump back then. I bought a FlowVis, but never got around to getting it installed.
Hey, were just chattin' and exploring ideas for the OP to consider. No worries. The flow rates used would be unique to each pool. I optimized mine with the FlowVis, based on manufacturers' requirement for each device. It's 40GPM for the solar, the vac is less, but I don't have the flow rate for that in front of me. For my setup, it's 2200RPM for the solar and 1900RPM for the VAC. I was once told (by my numbskull of a pool guy) that my NG heater would also require a good flow, pretty much the same as my solar. So if that's true, and I suspect it is as the heat exchange principles should be similar, there's no con there for the solar. (Sorry, too lazy to look up what my NG heater actually needs for flow, because I never use it.) They don't heat the same (BTU output), nor do they cost the same ($ per BTU). I just know my solar "costs" me what it would cost me for six 100W light bulbs, to raise my pool temp about 10° a day. I'm gonna say that's a tad less than what NG costs! So no con! 
I am curious about this too, do you run the pump for all daylight hours, or do you somehow get the pump to turn on when the sun is out? My system had a temperature sensor on the roof, and a temperature sensor in the pipes to check the pool water temp, if roof temp was sufficiently greater than water temp, then a valve actuator would open a valve to the panels and heat the pool until the desired water temp was achieved. Because I didn't always know when it would be sunny, I ran the pump for most of the daylight hours (on a few hours after sunrise, and off 1-1.5 hours before sunset)-- the temperature sensors wouldn't turn the pump on, just the valve, the pump had to already be running. Would be kinda cool to be able to run the pump more optimally timed to when its actually needed.
My runtime hours are flexible. I run solar around six hours a day. More if my daughter wants to swim, off if I know I won't be in the pool. My automation handles all that easily and I can control that on the fly, manually, or automatically from various devices (phone, computer, indoor panel, etc). Just like your system, mine knows when the panels are warm enough to heat my pool and when they're not. What might be different: my system ramps up the pump and engages the panels accordingly, until my target temp is reached. They can work without the sun, but better with, of course. But when the panels are not needed, my pump slows down to whatever speed I want for normal filtering. And I can program during what hours the solar can engage, so that's independent of when the filter mode is running. Is that answering your question? I think the automation and VS pump are solving for me what you were missing?
And just to throw another wrench in the works, I've added PV solar to my home, sized to accommodate my pool and its solar heater, so I actually have none of the concerns for energy efficiency I once did, or others might. I run my pool filter and vac and solar to optimize my experience (warm, clean, clear, etc) without regard for cost. Bragging, yes, but the point is: having PV solar in these times of ever-increasing electricity costs is a no brainer, IMO, and makes a lot of this nit-picking about energy efficiency somewhat (a lot) less of an issue. Both my solar systems will pay for themselves in short order, like five years or so, and then I'll be running and heating my pool for free. So that's a pro for solar, because there's not really a similar available solution for NG heating.
Hope this helps
