The only thing holding the button cap in is the switch itself. It works mostly fine, but if you get a cherry switch or an HBFS that isn’t up to spec then they can pop out. I’ve only had one of my HBSFs do this and I’ve installed probably 18 of them. Still, a strange design choice.
Edit for truth: I am a gamerfinger fanboy. I like the option to use the whole range of cherry switches. I don’t have many sticks that are ‘stock’ so the ability to get creative with buttons as well is fun.
I see you. You’re clearly bout that life 
The one issue I see happening is eventually someone going to have a button plunger come flying out and the stem for the switch is going to come out with it.
Leaving you with a busted MX style switch. It happened a few times to people on Geek Hack, and those guys love their keyboards.
I am thinking it first happen to one of the knock off razer switches first, but people been using (and abusing) the Cherry MX switches longer.
On that front I wonder how well Gateron switches works in a Game finger button?
Does anyone have experience with the HBFS-30-G2 + Razer (Orange/Green) switch combo? In keyboards the travel distance is reduced. I’m very curious if the same happens for the HBFS-30-G2…
Travel distance is 100% related to the switch itself, not the buttons. So if the engage point is shorter, then you will have a shorter travel required for engage.
Just to offer a dissenting opinion, I have a few sets of the HBFS/gamerfinger buttons, and I really like them and will continue to buy them. There are a few adjustments I would make on the 24mms, but overall they’re terrific buttons. If you want them to be totally silent you really need to use the red switches though. The blues are still more quiet than Sanwas.
100% ? close but not 100. Button design is important.
Proof: US/Euro style buttons (IL, Happ and clones) . If you press to much on the edge and on the opposite side of which the microswitch pin is located (the button plunger has two “legs” , one of which only contacts the microswitch pin, the other one presses into nothing) , which means that on one side you need to depress the button more than on the other, to obtain switch engagement, since the plunger tilts at an angle if you press on the wrong side. Actual travel of the leg that contacts the microswitch remains the same but you add the pre-travel of the opposite plunger leg, which means these type of buttons have a “sweet spot” that actuates before the diametrically opposite spot, generally you can feel it, it’s the stiffest zone on the edge of the button which is hardest to depress since the switch is right under it.
What you say is true only if there are zero tolerance issues for the plunger and if there is sufficient load on the button
Nendo is right, preload on the microswitch from a modified plunger can make the switch actuate with smaller travel .
The opposite would be a poor tolerance plunger whose bottom edge would not contact the base of the microswitch pin, it would need the travel from the plunger to contact the pin + actual travel of the switch pin into the said switch until actuation is reached. generally jap style buttons have acceptable tolerances.
I had the same idea as Nendo and you can use a shim made of thin enough plastic sheet (thickness = amount of travel that will be subtracted ) inside the plunger cavity that fits the microswitch pin.
Use it for a RG sanwa and presto you reduced the engage/actuation distance: all you need is a caliper and experiment with various thicknesses, placing the assembled button between the caliper jaws and paying attention when the assembly really actuates (use a battery and a LED for silent switches), pressing progressively, very slowly .
I tried it years ago with a seimitsu switch and it did work , but I found it too sensitive for me at the time.
Excessive preload on microswitches may reduce service time of switches that engage high ( too much wear and unnecessary high sensitivity) and may trigger unwanted actuation even with a shock to the panel, if you experiment with shims, you need to keep it real.