SpaceX does have a smallsat ride sharing program. The Starlink-8 launch was the first one (with 2 smallsats from SkySat), and the upcoming Starlink-9 also has 1 (BlackSky Global).
> If Spacex entered the market I think it would be game over
Why, though? The advantage of small launchers is the ability to put your smallsat into a dedicated orbit, which is not possible with a rideshare.
It's basically the choice of investing in R&D and adding launch mass for an on-sat propulsion system or paying extra for an exclusive ride.
An additional propulsion system not only increases risk, it also adds a substantial amount of cost and may turn a mission into the next bigger class - in terms of mass, overall cost, and development time - which may negate any saving in launch cost.
Edit: you can of course simply wait for an opportunity to launch as a secondary payload into the desired orbit, but that waiting time might be years...
The expectation is that for many smallsats the particular orbit does not matter, but cost matters a lot - so if SpaceX (which has much lower launch cost / kg) would get into this business, then people might be able to launch smallsats much cheaper.
But "many" isn't all - it might or might not be the majority.
In any case, the expected growth in the smallsat market is significant and the market might exceed the "traditional" satellite market by 2025 [1].
This leaves plenty of room for "boutique" launch services, especially considering that SpaceX will be their own best customer with their StarLink launch campaign.
So while they might be cheaper as far as $/kg is concerned, other factors (launch window, target orbit, payload integration, logistics, ...) have to be considered as well and the bigger the market, the more room for competition.
I'm not saying that no one will use (SpaceX) rideshare services - I'm just pointing out that SpaceX won't be the only reasonable option since $/kg isn't the only relevant metric to consider.
If Spacex entered the market I think it would be game over but they seem content to ignore it and focus on bigger payloads.
Last time I looked was blown away by the costs for a non-commercial 0.5U tubesat launch. $12,000 including all the hardware:
https://www.interorbital.com/Tubesat%20Kits