It will absolutely be used more often to exploit than to secure
Red is typically 'mischief' and Blue is typically 'mischief prevention.'
LotL is too advanced for script kiddies to pull off. You’re already dealing with an adversary with some sophistication, who likely have the time, skill, and resources to create their own LOLbins (or buy them on darknet forums). Infosec is an asymmetric conflict: the adversary has a much easier job IMO. It’s always easier to break something than to build it.
Blue/purple teams have little incentive to build their own LOLbins. There is another asymmetry here: blue teams are developing these to target only their networks. Red teams are developing them to target every network. Open source is the way to resolve that asymmetry.
With these becoming public, blue teams can simulate a LotL attack, develop indicators of attack, and write rules to trigger alerts.
In general, adversaries prefer their tricks to remain trade secrets. Once they’re known, documented, and commodified, they become far less useful.
Sure it might make it easier for red teams to help secure system, but for organizations bad at security they will be hit by that low hanging fruit methods, even if it might make more aware organizations more secure. Just because it lowers the point of entry.
(inb4 the comments, Squirrel itself attempts to strike a balance between usability and security, running only as the current user without admin limits its potential to be exploited, since any "I can hack Squirrel to run my code" trick is "Rather Involved Being On The Same Side Of The Airtight Hatch", as Raymond Chen would say)
That is pretty correct, but it is not necessarily an exploit or vulnerability in the binary. More often than not, it is a quirk or a way to use the binary which is unknown/uncommon, but might not appear on a defenders' radar.
We generally try to (ab)use functionality in pre-existing software to avoid security mechanisms (like AppLocker) and detections (like AV/EDR, or rules created by a SOC when analysing execution logs in a SIEM). Often we discover that a target computer has been hardened in some form or fashion, and we have to get "creative" when trying to download, execute or exfiltrate data during security assessments.