How effective can commercial-grade drones actually be?
Fundamentally, it feels like there would be a lot of issues around converting a commercial drone into a weapon.
First, navigation. GPS can be spoofed, downgraded or even disabled altogether, and is no real use against moving targets. Remote steering is possible, but can be jammed and is subject to lag - and unless you have some form of swarming technology, you need one warm body per drone. In theory, you could give a drone optical sensors and image recognition technology, or even have someone plant a beacon for a drone to home in on, but even those are subject to spoofing and jamming.
Then, there's range. A quick glance at Google indicates that commercial drone control range generally tops out at around 1km - though there is mention of one with a 7km range. Equally, flight time tops out at 15-30 minutes. I'd guess that this gas-turbine fuelled beastie may be capable of much more, but even then, it's unlikely to compare to a dedicated missile platform.
And there's more: commercial drones aren't designed to have a low radar profile, they're not designed for flying in adverse weather, they don't have systems redundancy, they don't have any defensive capabilities, they only have short range collision avoidance sensors, they're unarmored, they're not EMP hardened, etc.
Some of these things can be addressed as technology improves, or even retro-fitted on, but fundamentally, there's significant differences between the capabilities required for military and commercial activities.
Personally I think commercial-grade drones will take on a much smaller scale role - e.g. acting as disposable, short range scouts during infantry actions - if you're not too worried about making your presence known, they're perfect for city clearances and building searches. It's even possible that they could be used as beacons for their larger and more heavily armed military cousins...