Im going to break up your question into two distinct parts: why the PIRA suffered "relatively" few losses - my editorializing, not yours - as compared to other major militant groups, such as those in the American War on Terror, and why such a disparity exists between the number of security forces deaths in the Troubles relative to other European conflicts.

I want to tackle the second question first, oddly enough, because I think the premise is flawed. The Basque conflict is a good example. While you are correct that the number of deaths is not proportionally equivalent to the Troubles, I don't think "roughly equal" accurately describes the situation between losses on either side. Around 480+ members of what we might call "security forces" - mainly police, but also Spanish military - were killed from the period of 1959 - 2011. These numbers are pretty well established and can be found within the Spanish Minister of the Interior's conflict documentation. Deaths on the Basque nationalist side are harder to qualify; there are a range of estimates depending on who you ask; more importantly, the causes of death are often disputed. Many ETA members were killed by rival paramilitaries or during botched bombings. But Dissident, I know you're saying, didn't the PIRA also have rival paramilitaries? Didn't they also fall victim to premature detonations? Sure, and I'll come to that: this little aside serves to illustrate that sources focused on ETA deaths generally claim 100-150 deaths at the hands of the Spanish state; that's sill a disparity of almost 4:1.

Another conflict which actually involved the British military and saw a similar insurgency to Northern Ireland happened in Cyprus during the 1950s. EOKA militants lost around 80-100 people, depending on the sources, while the British military lost over 400. Again, this example is illustrative: low-intensity conflicts - especially internal ones - often see this type of disparity. And Cyprus was instructive both to the British military and the PIRA: EOKA prisoners were held alongside IRA members prior to the Border Campaign, and future members of the Provisionals communicated with them.

So, to turn to the PIRA: how in the world did they manage to attack the British military so "successfully"? Well, part of it was strategic. While attacks on Northern Irish police (the RUC) were common, during the late 1970s and 1980s the British military became the ideological focus of PIRA violence; the Brits were seen as an occupying force. However, their attacks are not the type many envision: while there were certainly a handful of protracted gun battles, many British deaths came from asymmetric "warfare" (or, depending on the politicization of your source, extreme "criminality). The most visible examples included: shooting off-duty soldiers who found themselves in the wrong place at the wrong time; bombing military parades; bombing pubs and discos that soldiers frequented. Provided your bomb or pistol worked, these operations exposed the fewest members of an IRA cell for the maximum casualty count. This is not to say, again, that more "traditional" forms of guerilla warfare never occurred. Roadside ambushes in rural counties and rocket attacks on helicopters/military installations were frequent in County Armagh and County Down. The Warrenpoint Ambush in 1979 was one example, and caused the largest losses seen by the British military during the entirety of Operation Banner: two strategically placed bombs were accompanied by rifle fire, and when the dust settled 18 soldiers had perished. These types of attacks add up, and help explain why the British lost so many soldiers to Republican violence.

Now, to turn to the first question. Thanks to the amazing work of resources like CAIN and journalists like David McKittrick, the death toll on the Republican militant side is very well established, unlike the conflict in Basque. Number of deaths and cause of death were pretty meticulously recorded during the conflict. So let's start with this: why did the British not kill more Irish Republicans? One easy answer is that the relative strength of the PIRA meant that numbers proportional to British losses were a statistical improbability, if not outright impossibility. Though everyone argues over the number of active PIRA volunteers (Tim Pat Coogan, Richard English, Daniel Finn, and the CAIN resource all have different numbers), it's likely that at its height the PIRA had around 1,000 to 1,500 "members". However, not all of these "members" were actively engaged in committing violence against the British military. Many were logistical support; the number of active Volunteers in PIRA Brigades was likely far, far lower than those numbers would have you believe. And given the type of low-profile actions they committed, it makes sense that the British military could not respond tit-for-tat. Moreover, when compared to operations in the War on Terror, the type of equipment and tactical opportunities differ wildly: you can surround Fallujah and have a protracted street campaign against insurgents, but you aren't lining up tanks around Belfast to root out members of insular communities. A far more effective tactic was infiltration and turning PIRA members into informants: the organization was so permeated with security leaks that they changed to a cell structure in the late 1970s, and even that did not help. Much of the damage inflicted on the PIRA was through the arrest of members owing to these breaches. So force size and the tactical limitations presented by policing an internal domestic conflict explain the smaller PIRA death count, along with other means of suppressing the insurrection like mass-arrests.

I can close by clearing up a few misconceptions/supporting a couple of your other hypotheses/finally returning to my caveat about ETA. I want to resist white-washing or demonizing the British military; their rules of engagement were usually strict, but miscommunication, over-zealous counter-terrorism ideologies, and illegal tactics all played a role in the conflict, even if they didn't wildly inflate the death count. The British killed innocent civilians on Bloody Sunday, an act for which no soldier has faced justice; undercover units like the Military Reaction Force and Force Research Unit played the paramilitaries against each other, and provided intelligence to competing groups to sow discord; mass arrests and internment of Northern Irish citizens occurred with the flimsiest of evidential justification. This is all to say that extreme even-handedness on the side of the British military was not the defining factor in the discrepancy you pointed out. By contrast, your point on the PIRA being particularly well-trained is bang-on: their Volunteers had some of the best improvised explosive makers going at the time, and while botched bombings occurred, this expertise helped keep those numbers down. Finally, returning to the Basque point, the nature of inter-paramilitary violence is vastly different in Northern Ireland. The right-wing groups fighting ETA were anti-leftist and anti-ETA: they targeted that group religiously. While Loyalist paramilitaries certainly attacked the PIRA as such, the Troubles descended into far greater sectarian retaliation by the 1980s, with bombings of pubs/shops/streets aimed at different ethnic/religious groups, as opposed to specific assaults on paramilitary strongholds. These attacks were targeted, yes, and they did kill members of militant groups: however, the nature of the violence looks rather different from that which occurred in the Basque Country.

Let's get you some sources.

For death totals in the Troubles:

CAIN Online Resource

"Lost Lives" by David McKittrick

For relative strengths of the PIRA:

See CAIN again

"The IRA" (Revised Edition) and "The Troubles" by Tim Pat Coogan

"Armed Struggle" by Richard English

"One Man's Terrorist: A Political History of the IRA" by Daniel Finn

For losses in the Basque conflict:

See the Spanish Government's Department of the Interior Ministry figures

Basque Eusko News Source

For general overview on the claims about PIRA tactics and the events that occured, also add:

"Making Sense of the Troubles" by David McKittrick

"A Secret History of the IRA" (2nd Edition) by Ed Moloney