> I even asked a few strangers what network they use and how come they have reception
However this is selection bias, you may actually have switched to a less popular network (which is better), or it could be coverage, it's hard to tell. The funny thing is that you can see how over time people will be switching back and forth between providers, even though there may actually be little change of coverage, any improvements could be more to do with movement of subscribers. Unfortunately 5G wont entirely eliminate this phenomenon because higher frequency comes with the disadvantage of being less penetrating to solid structures, so e.g in a train, or building people will sill be contending with each other over 4G quite often.
In the end, all that matters is what works, and you can only really test that empirically or by using other's to gauge performance as you did. Luckily you can now test 2 of the major UK networks without a contract, "Three" and "Vodafone" provide contract-less unlimmited month-to-month sim only through their sub-brands "Smarty" and "Voxi"... and I highly recommend using these over the main brand because you always have the flexibility to change quickly as network conditions change (I've switched between these two multiple times for a home LTE router due to changing user contention - which I've only inferred from the fact that there were no new towers in the area while maintaining the same RSRQ, and connecting to the same cell ids).
www.cellmapper.net is a good resource for roughly gauging coverage... but far better than any of the vague maps given on ISP sites, it actually shows you individual cells for each tower with rough directional areas, I've found it quite useful for debugging weird issues with antenna positioning on permanent LTE router setup due to locking onto bad towers (which unfortunately is not directly controllable on standard user equipment modems).
