Things I would have done differently:
1. I would have recognised the advantages of getting way more bays. More bays may seem more expensive but once you factor in the lower price of multiple lower capacity HDDs you can achieve a comparable price for a 'whole storage' system (especially when recycling existing smaller drives) so it can make financial sense. More spindles make for more speed and more flexibility, especially with Synology's SHR.
2. I would have thought more about form-factor. The DS units are nice looking, fractionally quieter and cheaper too but in reality they are devices that don't need physical interaction or to be seen. DS units are oddly shaped and don't really stack with other kit. I moved on to the RS short-depth rack-mounted format and (even without a rack) they are easier to stack with other kit and hide away somewhere. Getting everything into a rack, even at a later date, makes management simple.
3. Upgradability - something I did get right by getting a model with a PCIe slot. Didn't think I would need it but a change of circumstances and home-working made 10GbE networking suddenly an essential feature, along with an SSD-only volume.
4. Partner equipment - loads of things will come to mind when buying a NAS but arguably the best thing I bought for my NAS actually a small, cheap and low-power second NAS. Somewhere to locally back-up to, spends most of its time off-line for that virtual air-gap and has genuinely saved my bacon. Quite frankly it has turned DSM updates, major configuration changes, rebuilds, upgrades, drive issues and alike into a stress-free activity knowing that I have another local backup readily available.
5. Finally I would have asked an open question like yours on a forum such as this. I think it would have helped as in reality these units are not really NASes but small dedicated servers for home & work tasks alike. Something that is obvious to you already but back in the day I just thought about the storage, which is now the thing I think about the least.