Maybe you can use some of the learnings we had in our journey. When we started I wanted the whole thing wrt automation when we started our new-build project in the Netherlands. We started with the idea for loxone to "own" it all.... so we went with all loxone relais, tree touch buttons, code-touch, some air devices etc. Movement sensors basic 24V, dimmers (all 230V) where DMX based, Heating manifolds 230V with loxone extension (no longer sold) to control them. Doorbird for doorbell. Ventilation controlled through 0-10V. Hikvision camera's around the house. Basically use loxone wherever we could justify it... some thigns where just not right for us, so we went with generic components.
This was ~3 years ago... grand ideas for lawn robot, smart chargind of battery systems and other things one can think of...but first we had to move in. What happened since then.
1) after a few months we found the DMX solution not reliable enough, so switched all my DMX dimmers for Theben KNX dimmers.
2) Had space constraints in my wiring closet/space. So switched the Loxone Relais for Theben KNX relais (much smaller).
3) found out the hard way that there's such a large amount of steel in our concrete floors that air devices are useless. got rid of them and worked out a solution based on Shelly (Shelly H&T), this worked since I had done proper planning fro my network (Wifi) with UBNT unifi all over the place (and even outdoors).
4) Loxone discontinued the loxone extension, this means that when our installed extensions would ever fail I needed a plan to replace them given that I had no spares around, thus if some day I needed to get rid of them then any day would be a good day for this, so the month they announced these as EOS/EOL, I decided to swap them out for MDT KNX heating aktuators and a Weinzierl multi-IO 570 for 24V DI/DO contacts. At this point I still used the KNX devices as relatively dumb devices under control of loxone and with the ETS demo software.... just needed to manage a few ETS projects to work around the 5 device limit.
5) figured out that our ventilation system (Zehnder Comfoair Q600) has a nice app, and this is fine for me the few times a year I use it... reason is that the internal sensors of the ventilation system work perfect for our case, so no need for fancy control logic etc. if we ever have/want to do this then we can add a KNX gateway to it and integrate things that way. with the introduction of this app, I never used the 0-10V for our ventilation control again... whcih made me decide to get rid of it alltogether last month.
6) found out that the main reason loxone systems crash/die is because of bad SD cards... mostly because of frequent writes. therefore I decided to minimize the writes for statistics, and move all my statistics to our synology NAS with a few docker containers for IoBroker/influxDB/Grafana. I really like this, since it provide me the flexibility to also integrate directly with my shelly devices (Shelly 3EM), PV, Heatpump etc. and get all the statistics in one place... allowing me to build any dashboard I want. and as a bonus I can backup my statistics properly/easily.
7) our heatpump controller is smart enough to do the right thing, it also controls the heating/cooling pumps etc.... all it needs to know is if anywhere in the house there's heating/cooling demand. a few simple relais contacts do the trick. Oh, and our heating installer is not our electrician... they will not support the stuff the other guy did. So keeping these things separated using a simple (dry contact) interface makes this work out too...
8) my better half started to complain about not liking the loxone touch buttons... and the visitors all wondered how these things worked... Also Loxone started to act in ways which I don't like (no direct purchases or cases for end-users anymore, all through partners. Strange products which make me wonder who would ever want those... etc. The general feeling for loxone chnged from my side because of this). So we decided to go to relatively basic KNX switches (these are available nowadays, weren't there 4 years ago when we had to decide). Basic means, looks and feels like old fashioned buttons, include a temperature sensor, a led for visual feedback/alarm, and have enough capability to switch/dim the lights/blinds for a room... (we ended up with basic KNX buttons from Gira, since we had Gira material all around already, so color and finish matches). This was also the moment that I Invested in the BAB-tec app-module with ETS inside, and we moved all our movement sensors to KNX movement sensors. Took a few evenings of room-by-room migrations, some learning of ETS inside (once you "get it", it's really easy to work with) and some fine tuning... I now have the BAB-tec module as a plan-B in case I need to move away from loxone ever...
I now have decided that for heating/cooling, lighting and ventilation of the house I don't want to be depending on loxone. the only "large risk" I have for that is my KNX power supply, for which I have a spare one around (Which is also relatively cheap). To rach this, I only need to migrate my thermostat functionality from my loxone config to KNX... I know how to do this, just have to find a free day to do this... preferably not in the middle of the winter ;)
I still like loxone config A LOT, I also like the loxone app... so I keep the app integrated to KNX to allow control of the lights and to set a new target temperature for a room, just not to own the control...but mostly finetune if I ever need to (remember the nice thing about ETS inside is that you also configure this from your mobile device... so for real changes I can do these in a few clicks in KNX too, no need to startup the laptop). I also keep loxone around for the irrigation control, alarm of the house (with the code touch), switching outdoor lights based on location/timezone. it's worth the investment for just this alone (just check what an alarm system costs, and an irrigation computer and a KNX logic module... with some half decent UI).
We now only have a loxone Gen1 server with a tree extension and a code touch left that are loxone specific. I wouldn't mind moving to Gen2 some day, to keep loxone config working. One thing I'm toying with is whether I can make the loxone miniserver talk to KNX using UDP/HTTP in some way (either through the BAB-tec module with some application, or through a weinzierl IP Baos...) instead of through a KNX extension (I think it's overpriced). that's a project I want to figure out in the coming winter...
Also for charging of cars and batteries I've deceided to not rely on Loxone but make it work through independent solutions (likely a Zappi for the car). Again, this would become core to my house and thus should not rely on a single company.... same for lawn mowers, they'll be independent with their own sensors, it's afordable nowadays and will work fine together.
some of these things may apply to you, some may not.I do travel a fair bit for my work (just not last corona year), and things jsut need to work... for sure the house needs to keep "mostly working" if one component fails... especially when I'm not around for a week or so.
One thing I've learned is that the smart home market is developing really fast, some things are here to stay, some will pass by... and some are just gimmicks. try to build the things that need to last really well, the things that pass by or are gimmicks can be done in any other way (e.g. iobroker). I've also learned that integration should not be confused with "all in a single device/module/system"... integration is fine through APIs, interfaces and triggers. Also ask yourself the question what reall, really, really needs to keep on working in your house if things go sour with one vendor....Lastly I've learned that any verndor/solution you pick may develop in another direction then where you develop, e.g. loxone develops in the direction of installers/integrators which makes perfect sense for them to scale up/out as a company, it just doesn't fit too well with the open standards world and self-build/installers out there...
sorry for the longs (nto too well structured) e-mail. but I hope it helps you... long story short: integration isn't always the answer, integrate only where you have to...
Hidde