First you say that not showing security footage is “very suspicious,” but here you provide a reasonable, non-suspicious reason why the hotel might not want to send the footage to the 6:00 news. So which is it, suspicious or understandable?
If you’re going to propose some outlandish theories about the shooting, the onus is on you to provide evidence, not on everyone else to “provide fairly conclusive evidence there wasn’t.”
That’s an incredibly myopic way to view the overall cost of such sensors.
You have to consider the cost of wiring 3000+ sensors in to a central location. Ever seen what a bundle of 3000 wires looks like? Now imagine running that up and down a 480’ building. That’s a shit-ton of labor and more materials. Maybe they could be POE devices and tap into the hotel’s LAN, but you sure as hell aren’t going to get something like that for $6.50 at Home Depot.
Furthermore, the sensors you linked wouldn’t even be adequate for a hotel setting since they can’t be remotely monitored and would be trivially defeated by anyone with access to the area.
Finally any sensor has an inherent recurring maintenance and monitoring costs. Ultimately it’s something that will require man hours on a regular basis.
So now your 20k worth of sensors has ballooned into more like 2 million added cost for the construction of a hotel and some additional recurring costs.
Frequently, since I used to work for a telephone company.
That really is not how networks are built. Four wires running through each sensor should do it.
…all to deal with windows that get broken…how often? 
Seriously, I have no idea how frequently windows get broken in a place like Mandalay bay.
Moreover, why would the hotel need an alarm to learn about a broken window in a guest’s room? In 99.99999999% of instances, the window will have been broken by accident, and the room’s occupant will be calling the front desk promptly to inform them while simultaneously staying safely away from the now-unprotected ledge.
Which means that the sole purpose of this entire alarm system would be to inform the front desk that Stephen Paddock is about to shoot up a concert 1 block north and across the street that has nothing to do with Mandalay Bay. So how does Mandalay Bay benefit from installing such a system? What’s their ROI?
But wait, there’s more. If you want to be thorough about this, then every casino/hotel on the strip should be fitted with this system. At which point we’re talking about tens of millions of dollars, all to get an early alert that Stephen Paddock is about to start shooting from Mandalay Bay. Seriously, this incident has happened exactly once in the history of the strip. The cost of a truly effective early warning system is not justified.
Did you work for free?
You’ll notice that I said POE (power over ethernet) devices could alternatively be used which could then piggyback on existing/already-planned network infrastructure. But those would be way more than $6.50 a pop. And that failing, the only way to be able to precisely locate a break of an individual window would be to have separate signal leads for each. That is how electricity works. Now certainly you could coalesce these signal leads into smaller bundles, perhaps 1 per floor, to a lightweight server and then pipe that metadata to a centralized server, but again … labor … materials cost … maintenance.
The bottom line is, regardless of how you go about skinning this cat, the cost of the actual sensors is barely the tip of the iceberg.